publication_date
stringlengths 20
20
| doc_url
stringlengths 33
34
| id
stringlengths 33
34
| authors
stringlengths 5
20.1k
| update_date
stringlengths 20
20
| category_all
stringlengths 5
138
| abstract
stringlengths 123
3.09k
| category_primary
stringclasses 78
values | title
stringlengths 10
211
|
|---|---|---|---|---|---|---|---|---|
2023-06-26T08:48:08Z | http://arxiv.org/pdf/2306.14517v1 | http://arxiv.org/abs/2306.14517v1 | Samuel Cahyawijaya, Holy Lovenia, Willy Chung, Rita Frieske, Zihan Liu, Pascale Fung | 2023-06-26T08:48:08Z | cs.CL, cs.SD, eess.AS | Speech emotion recognition plays a crucial role in human-computer
interactions. However, most speech emotion recognition research is biased
toward English-speaking adults, which hinders its applicability to other
demographic groups in different languages and age groups. In this work, we
analyze the transferability of emotion recognition across three different
languages--English, Mandarin Chinese, and Cantonese; and 2 different age
groups--adults and the elderly. To conduct the experiment, we develop an
English-Mandarin speech emotion benchmark for adults and the elderly, BiMotion,
and a Cantonese speech emotion dataset, YueMotion. This study concludes that
different language and age groups require specific speech features, thus making
cross-lingual inference an unsuitable method. However, cross-group data
augmentation is still beneficial to regularize the model, with linguistic
distance being a significant influence on cross-lingual transferability. We
release publicly release our code at https://github.com/HLTCHKUST/elderly_ser. | cs.CL | Cross-Lingual Cross-Age Group Adaptation for Low-Resource Elderly Speech Emotion Recognition |
2023-06-26T08:45:47Z | http://arxiv.org/pdf/2306.14514v2 | http://arxiv.org/abs/2306.14514v2 | Seugnjun Lee, Hyeonseok Moon, Chanjun Park, Heuiseok Lim | 2023-06-27T06:59:28Z | cs.CL, cs.AI | In this paper, we introduce a data-driven approach for Formality-Sensitive
Machine Translation (FSMT) that caters to the unique linguistic properties of
four target languages. Our methodology centers on two core strategies: 1)
language-specific data handling, and 2) synthetic data generation using
large-scale language models and empirical prompt engineering. This approach
demonstrates a considerable improvement over the baseline, highlighting the
effectiveness of data-centric techniques. Our prompt engineering strategy
further improves performance by producing superior synthetic translation
examples. | cs.CL | Data-Driven Approach for Formality-Sensitive Machine Translation: Language-Specific Handling and Synthetic Data Generation |
2023-06-26T07:23:47Z | http://arxiv.org/pdf/2306.14470v1 | http://arxiv.org/abs/2306.14470v1 | Dahyun Jung, Jaehyung Seo, Jaewook Lee, Chanjun Park, Heuiseok Lim | 2023-06-26T07:23:47Z | cs.CL, cs.AI | Generative commonsense reasoning refers to the task of generating acceptable
and logical assumptions about everyday situations based on commonsense
understanding. By utilizing an existing dataset such as Korean CommonGen,
language generation models can learn commonsense reasoning specific to the
Korean language. However, language models often fail to consider the
relationships between concepts and the deep knowledge inherent to concepts. To
address these limitations, we propose a method to utilize the Korean knowledge
graph data for text generation. Our experimental result shows that the proposed
method can enhance the efficiency of Korean commonsense inference, thereby
underlining the significance of employing supplementary data. | cs.CL | Knowledge Graph-Augmented Korean Generative Commonsense Reasoning |
2023-06-26T07:00:38Z | http://arxiv.org/pdf/2306.14457v1 | http://arxiv.org/abs/2306.14457v1 | Andrea Bacciu, Giovanni Trappolini, Andrea Santilli, Emanuele Rodolà, Fabrizio Silvestri | 2023-06-26T07:00:38Z | cs.CL | This paper presents Fauno, the first and largest open-source Italian
conversational Large Language Model (LLM). Our goal with Fauno is to
democratize the study of LLMs in Italian, demonstrating that obtaining a
fine-tuned conversational bot with a single GPU is possible. In addition, we
release a collection of datasets for conversational AI in Italian. The datasets
on which we fine-tuned Fauno include various topics such as general question
answering, computer science, and medical questions. We release our code and
datasets on \url{https://github.com/RSTLess-research/Fauno-Italian-LLM} | cs.CL | Fauno: The Italian Large Language Model that will leave you senza parole! |
2023-06-26T05:04:58Z | http://arxiv.org/pdf/2306.14422v2 | http://arxiv.org/abs/2306.14422v2 | Wen-Chin Huang, Lester Phillip Violeta, Songxiang Liu, Jiatong Shi, Tomoki Toda | 2023-07-06T08:17:31Z | cs.SD, cs.CL, eess.AS | We present the latest iteration of the voice conversion challenge (VCC)
series, a bi-annual scientific event aiming to compare and understand different
voice conversion (VC) systems based on a common dataset. This year we shifted
our focus to singing voice conversion (SVC), thus named the challenge the
Singing Voice Conversion Challenge (SVCC). A new database was constructed for
two tasks, namely in-domain and cross-domain SVC. The challenge was run for two
months, and in total we received 26 submissions, including 2 baselines. Through
a large-scale crowd-sourced listening test, we observed that for both tasks,
although human-level naturalness was achieved by the top system, no team was
able to obtain a similarity score as high as the target speakers. Also, as
expected, cross-domain SVC is harder than in-domain SVC, especially in the
similarity aspect. We also investigated whether existing objective measurements
were able to predict perceptual performance, and found that only few of them
could reach a significant correlation. | cs.SD | The Singing Voice Conversion Challenge 2023 |
2023-06-26T03:06:57Z | http://arxiv.org/pdf/2306.14393v1 | http://arxiv.org/abs/2306.14393v1 | Junyan Li, Li Lyna Zhang, Jiahang Xu, Yujing Wang, Shaoguang Yan, Yunqing Xia, Yuqing Yang, Ting Cao, Hao Sun, Weiwei Deng, Qi Zhang, Mao Yang | 2023-06-26T03:06:57Z | cs.CL | Deploying pre-trained transformer models like BERT on downstream tasks in
resource-constrained scenarios is challenging due to their high inference cost,
which grows rapidly with input sequence length. In this work, we propose a
constraint-aware and ranking-distilled token pruning method ToP, which
selectively removes unnecessary tokens as input sequence passes through layers,
allowing the model to improve online inference speed while preserving accuracy.
ToP overcomes the limitation of inaccurate token importance ranking in the
conventional self-attention mechanism through a ranking-distilled token
distillation technique, which distills effective token rankings from the final
layer of unpruned models to early layers of pruned models. Then, ToP introduces
a coarse-to-fine pruning approach that automatically selects the optimal subset
of transformer layers and optimizes token pruning decisions within these layers
through improved $L_0$ regularization. Extensive experiments on GLUE benchmark
and SQuAD tasks demonstrate that ToP outperforms state-of-the-art token pruning
and model compression methods with improved accuracy and speedups. ToP reduces
the average FLOPs of BERT by 8.1x while achieving competitive accuracy on GLUE,
and provides a real latency speedup of up to 7.4x on an Intel CPU. | cs.CL | Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference |
2023-06-26T01:40:28Z | http://arxiv.org/pdf/2306.14377v1 | http://arxiv.org/abs/2306.14377v1 | Chanjun Park, Seonmin Koo, Seolhwa Lee, Jaehyung Seo, Sugyeong Eo, Hyeonseok Moon, Heuiseok Lim | 2023-06-26T01:40:28Z | cs.CL, cs.AI | Data-centric AI approach aims to enhance the model performance without
modifying the model and has been shown to impact model performance positively.
While recent attention has been given to data-centric AI based on synthetic
data, due to its potential for performance improvement, data-centric AI has
long been exclusively validated using real-world data and publicly available
benchmark datasets. In respect of this, data-centric AI still highly depends on
real-world data, and the verification of models using synthetic data has not
yet been thoroughly carried out. Given the challenges above, we ask the
question: Does data quality control (noise injection and balanced data), a
data-centric AI methodology acclaimed to have a positive impact, exhibit the
same positive impact in models trained solely with synthetic data? To address
this question, we conducted comparative analyses between models trained on
synthetic and real-world data based on grammatical error correction (GEC) task.
Our experimental results reveal that the data quality control method has a
positive impact on models trained with real-world data, as previously reported
in existing studies, while a negative impact is observed in models trained
solely on synthetic data. | cs.CL | Synthetic Alone: Exploring the Dark Side of Synthetic Data for Grammatical Error Correction |
2023-06-26T01:33:58Z | http://arxiv.org/pdf/2306.14374v1 | http://arxiv.org/abs/2306.14374v1 | Damrin Kim, NamHyeok Kim, Chanjun Park, Harksoo Kim | 2023-06-26T01:33:58Z | cs.CL, cs.AI | This paper presents a novel approach of leveraging Inter-Annotator Agreement
(IAA), traditionally used for assessing labeling consistency, to optimize Data
Management Operations (DMOps). We advocate for the use of IAA in predicting the
labeling quality of individual annotators, leading to cost and time efficiency
in data production. Additionally, our work highlights the potential of IAA in
forecasting document difficulty, thereby boosting the data construction
process's overall efficiency. This research underscores IAA's broader
application potential in data-driven research optimization and holds
significant implications for large-scale data projects prioritizing efficiency,
cost reduction, and high-quality data. | cs.CL | Transcending Traditional Boundaries: Leveraging Inter-Annotator Agreement (IAA) for Enhancing Data Management Operations (DMOps) |
2023-06-26T01:28:58Z | http://arxiv.org/pdf/2306.14373v1 | http://arxiv.org/abs/2306.14373v1 | NamHyeok Kim, Chanjun Park | 2023-06-26T01:28:58Z | cs.CL, cs.AI | Inter-Annotator Agreement (IAA) is commonly used as a measure of label
consistency in natural language processing tasks. However, in real-world
scenarios, IAA has various roles and implications beyond its traditional usage.
In this paper, we not only consider IAA as a measure of consistency but also as
a versatile tool that can be effectively utilized in practical applications.
Moreover, we discuss various considerations and potential concerns when
applying IAA and suggest strategies for effectively navigating these
challenges. | cs.CL | Inter-Annotator Agreement in the Wild: Uncovering Its Emerging Roles and Considerations in Real-World Scenarios |
2023-06-25T19:23:21Z | http://arxiv.org/pdf/2306.14321v1 | http://arxiv.org/abs/2306.14321v1 | Yilun Zhao, Chen Zhao, Linyong Nan, Zhenting Qi, Wenlin Zhang, Xiangru Tang, Boyu Mi, Dragomir Radev | 2023-06-25T19:23:21Z | cs.CL, cs.AI | Despite significant progress having been made in question answering on
tabular data (Table QA), it's unclear whether, and to what extent existing
Table QA models are robust to task-specific perturbations, e.g., replacing key
question entities or shuffling table columns. To systematically study the
robustness of Table QA models, we propose a benchmark called RobuT, which
builds upon existing Table QA datasets (WTQ, WikiSQL-Weak, and SQA) and
includes human-annotated adversarial perturbations in terms of table header,
table content, and question. Our results indicate that both state-of-the-art
Table QA models and large language models (e.g., GPT-3) with few-shot learning
falter in these adversarial sets. We propose to address this problem by using
large language models to generate adversarial examples to enhance training,
which significantly improves the robustness of Table QA models. Our data and
code is publicly available at https://github.com/yilunzhao/RobuT. | cs.CL | RobuT: A Systematic Study of Table QA Robustness Against Human-Annotated Adversarial Perturbations |
2023-06-25T18:48:21Z | http://arxiv.org/pdf/2306.14310v1 | http://arxiv.org/abs/2306.14310v1 | Jungbae Park, Seungtaek Choi | 2023-06-25T18:48:21Z | cs.CL, cs.SD, eess.AS | Integrating automatic speech scoring/assessment systems has become a critical
aspect of second-language speaking education. With self-supervised learning
advancements, end-to-end speech scoring approaches have exhibited promising
results. However, this study highlights the significant decrease in the
performance of speech scoring systems in new question contexts, thereby
identifying this as a cold start problem in terms of items. With the finding of
cold-start phenomena, this paper seeks to alleviate the problem by following
methods: 1) prompt embeddings, 2) question context embeddings using BERT or
CLIP models, and 3) choice of the pretrained acoustic model. Experiments are
conducted on TOEIC speaking test datasets collected from
English-as-a-second-language (ESL) learners rated by professional TOEIC
speaking evaluators. The results demonstrate that the proposed framework not
only exhibits robustness in a cold-start environment but also outperforms the
baselines for known content. | cs.CL | Addressing Cold Start Problem for End-to-end Automatic Speech Scoring |
2023-06-25T18:40:43Z | http://arxiv.org/pdf/2306.14308v1 | http://arxiv.org/abs/2306.14308v1 | Xiao Ma, Swaroop Mishra, Ahmad Beirami, Alex Beutel, Jilin Chen | 2023-06-25T18:40:43Z | cs.CL, cs.AI | Language models still struggle on moral reasoning, despite their impressive
performance in many other tasks. In particular, the Moral Scenarios task in
MMLU (Multi-task Language Understanding) is among the worst performing tasks
for many language models, including GPT-3. In this work, we propose a new
prompting framework, Thought Experiments, to teach language models to do better
moral reasoning using counterfactuals. Experiment results show that our
framework elicits counterfactual questions and answers from the model, which in
turn helps improve the accuracy on Moral Scenarios task by 9-16% compared to
other zero-shot baselines. Interestingly, unlike math reasoning tasks,
zero-shot Chain-of-Thought (CoT) reasoning doesn't work out of the box, and
even reduces accuracy by around 4% compared to direct zero-shot. We further
observed that with minimal human supervision in the form of 5 few-shot
examples, the accuracy of the task can be improved to as much as 80%. | cs.CL | Let's Do a Thought Experiment: Using Counterfactuals to Improve Moral Reasoning |
2023-06-25T15:09:21Z | http://arxiv.org/pdf/2306.14264v1 | http://arxiv.org/abs/2306.14264v1 | Jayesh Songara, Shivam Pande, Shabnam Choudhury, Biplab Banerjee, Rajbabu Velmurugan | 2023-06-25T15:09:21Z | cs.CV, cs.CL, cs.LG | In this research, we deal with the problem of visual question answering (VQA)
in remote sensing. While remotely sensed images contain information significant
for the task of identification and object detection, they pose a great
challenge in their processing because of high dimensionality, volume and
redundancy. Furthermore, processing image information jointly with language
features adds additional constraints, such as mapping the corresponding image
and language features. To handle this problem, we propose a cross attention
based approach combined with information maximization. The CNN-LSTM based
cross-attention highlights the information in the image and language modalities
and establishes a connection between the two, while information maximization
learns a low dimensional bottleneck layer, that has all the relevant
information required to carry out the VQA task. We evaluate our method on two
VQA remote sensing datasets of different resolutions. For the high resolution
dataset, we achieve an overall accuracy of 79.11% and 73.87% for the two test
sets while for the low resolution dataset, we achieve an overall accuracy of
85.98%. | cs.CV | Visual Question Answering in Remote Sensing with Cross-Attention and Multimodal Information Bottleneck |
2023-06-25T12:08:44Z | http://arxiv.org/pdf/2306.14222v2 | http://arxiv.org/abs/2306.14222v2 | Haohan Zhang, Fengrui Hua, Chengjin Xu, Hao Kong, Ruiting Zuo, Jian Guo | 2024-05-04T13:02:32Z | cs.CL, cs.AI, q-fin.ST | The rapid advancement of Large Language Models (LLMs) has spurred discussions
about their potential to enhance quantitative trading strategies. LLMs excel in
analyzing sentiments about listed companies from financial news, providing
critical insights for trading decisions. However, the performance of LLMs in
this task varies substantially due to their inherent characteristics. This
paper introduces a standardized experimental procedure for comprehensive
evaluations. We detail the methodology using three distinct LLMs, each
embodying a unique approach to performance enhancement, applied specifically to
the task of sentiment factor extraction from large volumes of Chinese news
summaries. Subsequently, we develop quantitative trading strategies using these
sentiment factors and conduct back-tests in realistic scenarios. Our results
will offer perspectives about the performances of Large Language Models applied
to extracting sentiments from Chinese news texts. | cs.CL | Unveiling the Potential of Sentiment: Can Large Language Models Predict Chinese Stock Price Movements? |
2023-06-25T10:54:24Z | http://arxiv.org/pdf/2306.14203v2 | http://arxiv.org/abs/2306.14203v2 | Shester Gueuwou, Rose-Mary Owusuaa Mensah Gyening | 2023-06-27T18:39:53Z | cs.CL | On December 7, 2020, Ghanaians participated in the polls to determine their
president for the next four years. To gain insights from this presidential
election, we conducted stance analysis (which is not always equivalent to
sentiment analysis) to understand how Twitter, a popular social media platform,
reflected the opinions of its users regarding the two main presidential
candidates. We collected a total of 99,356 tweets using the Twitter API
(Tweepy) and manually annotated 3,090 tweets into three classes: Against,
Neutral, and Support. We then performed preprocessing on the tweets. The
resulting dataset was evaluated using two lexicon-based approaches, VADER and
TextBlob, as well as five supervised machine learning-based approaches: Support
Vector Machine (SVM), Logistic Regression (LR), Multinomial Na\"ive Bayes
(MNB), Stochastic Gradient Descent (SGD), and Random Forest (RF), based on
metrics such as accuracy, precision, recall, and F1-score. The best performance
was achieved by Logistic Regression with an accuracy of 71.13%. We utilized
Logistic Regression to classify all the extracted tweets and subsequently
conducted an analysis and discussion of the results. For access to our data and
code, please visit:
https://github.com/ShesterG/Stance-Detection-Ghana-2020-Elections.git | cs.CL | Stance Prediction and Analysis of Twitter data : A case study of Ghana 2020 Presidential Elections |
2023-06-25T10:16:49Z | http://arxiv.org/pdf/2306.14192v3 | http://arxiv.org/abs/2306.14192v3 | Pamela Fleischmann, Jonas Höfer, Annika Huch, Dirk Nowotka | 2023-09-11T06:47:27Z | math.CO, cs.CL | In 1991 H\'ebrard introduced a factorization of words that turned out to be a
powerful tool for the investigation of a word's scattered factors (also known
as (scattered) subwords or subsequences). Based on this, first Karandikar and
Schnoebelen introduced the notion of $k$-richness and later on Barker et al.
the notion of $k$-universality. In 2022 Fleischmann et al. presented a
generalization of the arch factorization by intersecting the arch factorization
of a word and its reverse. While the authors merely used this factorization for
the investigation of shortest absent scattered factors, in this work we
investigate this new $\alpha$-$\beta$-factorization as such. We characterize
the famous Simon congruence of $k$-universal words in terms of $1$-universal
words. Moreover, we apply these results to binary words. In this special case,
we obtain a full characterization of the classes and calculate the index of the
congruence. Lastly, we start investigating the ternary case, present a full
list of possibilities for $\alpha\beta\alpha$-factors, and characterize their
congruence. | math.CO | $α$-$β$-Factorization and the Binary Case of Simon's Congruence |
2023-06-25T09:03:56Z | http://arxiv.org/pdf/2306.14176v1 | http://arxiv.org/abs/2306.14176v1 | Tongtao Ling, Lei Chen, Huangxu Sheng, Zicheng Cai, Hai-Lin Liu | 2023-06-25T09:03:56Z | cs.CL | The traditional way of sentence-level event detection involves two important
subtasks: trigger identification and trigger classifications, where the
identified event trigger words are used to classify event types from sentences.
However, trigger classification highly depends on abundant annotated trigger
words and the accuracy of trigger identification. In a real scenario,
annotating trigger words is time-consuming and laborious. For this reason, we
propose a trigger-free event detection model, which transforms event detection
into a two-tower model based on machine reading comprehension and prompt
learning. Compared to existing trigger-based and trigger-free methods,
experimental studies on two event detection benchmark datasets (ACE2005 and
MAVEN) have shown that the proposed approach can achieve competitive
performance. | cs.CL | Sentence-level Event Detection without Triggers via Prompt Learning and Machine Reading Comprehension |
2023-06-25T07:38:43Z | http://arxiv.org/pdf/2306.14152v1 | http://arxiv.org/abs/2306.14152v1 | Siyu Ren, Kenny Q. Zhu | 2023-06-25T07:38:43Z | cs.CL | The components underpinning PLMs -- large weight matrices -- were shown to
bear considerable redundancy. Matrix factorization, a well-established
technique from matrix theory, has been utilized to reduce the number of
parameters in PLM. However, it fails to retain satisfactory performance under
moderate to high compression rate. In this paper, we identify the
\textit{full-rankness} of fine-tuned PLM as the fundamental bottleneck for the
failure of matrix factorization and explore the use of network pruning to
extract low-rank sparsity pattern desirable to matrix factorization. We find
such low-rank sparsity pattern exclusively exists in models generated by
first-order pruning, which motivates us to unite the two approaches and achieve
more effective model compression. We further propose two techniques:
sparsity-aware SVD and mixed-rank fine-tuning, which improve the initialization
and training of the compression procedure, respectively. Experiments on GLUE
and question-answering tasks show that the proposed method has superior
compression-performance trade-off compared to existing approaches. | cs.CL | Low-Rank Prune-And-Factorize for Language Model Compression |
2023-06-25T07:25:14Z | http://arxiv.org/pdf/2306.14149v1 | http://arxiv.org/abs/2306.14149v1 | Xiao Zhang, Heqi Zheng, Yuxiang Nie, Heyan Huang, Xian-Ling Mao | 2023-06-25T07:25:14Z | cs.CL | Scientific machine reading comprehension (SMRC) aims to understand scientific
texts through interactions with humans by given questions. As far as we know,
there is only one dataset focused on exploring full-text scientific machine
reading comprehension. However, the dataset has ignored the fact that different
readers may have different levels of understanding of the text, and only
includes single-perspective question-answer pairs, leading to a lack of
consideration of different perspectives. To tackle the above problem, we
propose a novel multi-perspective SMRC dataset, called SciMRC, which includes
perspectives from beginners, students and experts. Our proposed SciMRC is
constructed from 741 scientific papers and 6,057 question-answer pairs. Each
perspective of beginners, students and experts contains 3,306, 1,800 and 951 QA
pairs, respectively. The extensive experiments on SciMRC by utilizing
pre-trained models suggest the importance of considering perspectives of SMRC,
and demonstrate its challenging nature for machine comprehension. | cs.CL | SciMRC: Multi-perspective Scientific Machine Reading Comprehension |
2023-06-25T06:46:36Z | http://arxiv.org/pdf/2306.14145v1 | http://arxiv.org/abs/2306.14145v1 | Sen Liu, Yiwei Guo, Chenpeng Du, Xie Chen, Kai Yu | 2023-06-25T06:46:36Z | cs.SD, cs.CL, eess.AS | Although high-fidelity speech can be obtained for intralingual speech
synthesis, cross-lingual text-to-speech (CTTS) is still far from satisfactory
as it is difficult to accurately retain the speaker timbres(i.e. speaker
similarity) and eliminate the accents from their first language(i.e.
nativeness). In this paper, we demonstrated that vector-quantized(VQ) acoustic
feature contains less speaker information than mel-spectrogram. Based on this
finding, we propose a novel dual speaker embedding TTS (DSE-TTS) framework for
CTTS with authentic speaking style. Here, one embedding is fed to the acoustic
model to learn the linguistic speaking style, while the other one is integrated
into the vocoder to mimic the target speaker's timbre. Experiments show that by
combining both embeddings, DSE-TTS significantly outperforms the
state-of-the-art SANE-TTS in cross-lingual synthesis, especially in terms of
nativeness. | cs.SD | DSE-TTS: Dual Speaker Embedding for Cross-Lingual Text-to-Speech |
2023-06-25T05:57:01Z | http://arxiv.org/pdf/2306.14135v1 | http://arxiv.org/abs/2306.14135v1 | Minxue Xia, Hao Zhu | 2023-06-25T05:57:01Z | cs.CL, cs.LG | Interpretability benefits the theoretical understanding of representations.
Existing word embeddings are generally dense representations. Hence, the
meaning of latent dimensions is difficult to interpret. This makes word
embeddings like a black-box and prevents them from being human-readable and
further manipulation. Many methods employ sparse representation to learn
interpretable word embeddings for better interpretability. However, they also
suffer from the unstable issue of grouped selection in $\ell1$ and online
dictionary learning. Therefore, they tend to yield different results each time.
To alleviate this challenge, we propose a novel method to associate data
self-representation with a shallow neural network to learn expressive,
interpretable word embeddings. In experiments, we report that the resulting
word embeddings achieve comparable and even slightly better interpretability
than baseline embeddings. Besides, we also evaluate that our approach performs
competitively well on all downstream tasks and outperforms benchmark embeddings
on a majority of them. | cs.CL | Interpretable Neural Embeddings with Sparse Self-Representation |
2023-06-25T04:33:56Z | http://arxiv.org/pdf/2306.14122v3 | http://arxiv.org/abs/2306.14122v3 | Feng Chen, Yujian Feng | 2023-08-23T05:04:58Z | cs.CL, cs.CV | Multimodal Named Entity Recognition (MNER) and Multimodal Relation Extraction
(MRE) necessitate the fundamental reasoning capacity for intricate linguistic
and multimodal comprehension. In this study, we explore distilling the
reasoning ability of large language models (LLMs) into a more compact student
model by generating a \textit{chain of thought} (CoT) -- a sequence of
intermediate reasoning steps. Specifically, we commence by exemplifying the
elicitation of such reasoning ability from LLMs through CoT prompts covering
multi-grain (noun, sentence, multimodality) and data-augmentation (style,
entity, image) dimensions. Subsequently, we present a novel conditional prompt
distillation method to assimilate the commonsense reasoning ability from LLMs,
thereby enhancing the utility of the student model in addressing text-only
inputs without the requisite addition of image and CoT knowledge. Extensive
experiments reveal that our approach attains state-of-the-art accuracy and
manifests a plethora of advantages concerning interpretability, data
efficiency, and cross-domain generalization on MNER and MRE datasets. | cs.CL | Chain-of-Thought Prompt Distillation for Multimodal Named Entity Recognition and Multimodal Relation Extraction |
2023-06-25T03:34:06Z | http://arxiv.org/pdf/2306.14115v2 | http://arxiv.org/abs/2306.14115v2 | Wenbo Zhang, Tong Wu, Yunlong Wang, Yong Cai, Hengrui Cai | 2023-09-09T03:50:04Z | cs.LG, cs.AI, cs.CL, stat.ME, stat.ML | With recent advances in natural language processing, rationalization becomes
an essential self-explaining diagram to disentangle the black box by selecting
a subset of input texts to account for the major variation in prediction. Yet,
existing association-based approaches on rationalization cannot identify true
rationales when two or more snippets are highly inter-correlated and thus
provide a similar contribution to prediction accuracy, so-called spuriousness.
To address this limitation, we novelly leverage two causal desiderata,
non-spuriousness and efficiency, into rationalization from the causal inference
perspective. We formally define a series of probabilities of causation based on
a newly proposed structural causal model of rationalization, with its
theoretical identification established as the main component of learning
necessary and sufficient rationales. The superior performance of the proposed
causal rationalization is demonstrated on real-world review and medical
datasets with extensive experiments compared to state-of-the-art methods. | cs.LG | Towards Trustworthy Explanation: On Causal Rationalization |
2023-06-25T02:24:30Z | http://arxiv.org/pdf/2306.14096v5 | http://arxiv.org/abs/2306.14096v5 | Yinyu Lan, Yanru Wu, Wang Xu, Weiqiang Feng, Youhao Zhang | 2023-09-15T08:19:44Z | cs.CL, cs.AI | Entity-level fine-grained sentiment analysis in the financial domain is a
crucial subtask of sentiment analysis and currently faces numerous challenges.
The primary challenge stems from the lack of high-quality and large-scale
annotated corpora specifically designed for financial text sentiment analysis,
which in turn limits the availability of data necessary for developing
effective text processing techniques. Recent advancements in large language
models (LLMs) have yielded remarkable performance in natural language
processing tasks, primarily centered around language pattern matching. In this
paper, we propose a novel and extensive Chinese fine-grained financial
sentiment analysis dataset, FinChina SA, for enterprise early warning. We
thoroughly evaluate and experiment with well-known existing open-source LLMs
using our dataset. We firmly believe that our dataset will serve as a valuable
resource to advance the exploration of real-world financial sentiment analysis
tasks, which should be the focus of future research. The FinChina SA dataset is
publicly available at https://github.com/YerayL/FinChina-SA | cs.CL | Chinese Fine-Grained Financial Sentiment Analysis with Large Language Models |
2023-06-24T22:00:06Z | http://arxiv.org/pdf/2306.14067v1 | http://arxiv.org/abs/2306.14067v1 | Michael Ogezi, Bradley Hauer, Talgat Omarov, Ning Shi, Grzegorz Kondrak | 2023-06-24T22:00:06Z | cs.CL | We describe the systems of the University of Alberta team for the
SemEval-2023 Visual Word Sense Disambiguation (V-WSD) Task. We present a novel
algorithm that leverages glosses retrieved from BabelNet, in combination with
text and image encoders. Furthermore, we compare language-specific encoders
against the application of English encoders to translated texts. As the
contexts given in the task datasets are extremely short, we also experiment
with augmenting these contexts with descriptions generated by a language model.
This yields substantial improvements in accuracy. We describe and evaluate
additional V-WSD methods which use image generation and text-conditioned image
segmentation. Overall, the results of our official submission rank us 18 out of
56 teams. Some of our unofficial results are even better than the official
ones. Our code is publicly available at https://github.com/UAlberta-NLP/v-wsd. | cs.CL | UAlberta at SemEval-2023 Task 1: Context Augmentation and Translation for Multilingual Visual Word Sense Disambiguation |
2023-06-24T21:05:02Z | http://arxiv.org/pdf/2306.14060v1 | http://arxiv.org/abs/2306.14060v1 | Liunian Harold Li, Zi-Yi Dou, Nanyun Peng, Kai-Wei Chang | 2023-06-24T21:05:02Z | cs.CV, cs.CL, cs.LG | Recent development in vision-language approaches has instigated a paradigm
shift in learning visual recognition models from language supervision. These
approaches align objects with language queries (e.g. "a photo of a cat") and
improve the models' adaptability to identify novel objects and domains.
Recently, several studies have attempted to query these models with complex
language expressions that include specifications of fine-grained semantic
details, such as attributes, shapes, textures, and relations. However, simply
incorporating language descriptions as queries does not guarantee accurate
interpretation by the models. In fact, our experiments show that GLIP, the
state-of-the-art vision-language model for object detection, often disregards
contextual information in the language descriptions and instead relies heavily
on detecting objects solely by their names. To tackle the challenges, we
propose a new description-conditioned (DesCo) paradigm of learning object
recognition models with rich language descriptions consisting of two major
innovations: 1) we employ a large language model as a commonsense knowledge
engine to generate rich language descriptions of objects based on object names
and the raw image-text caption; 2) we design context-sensitive queries to
improve the model's ability in deciphering intricate nuances embedded within
descriptions and enforce the model to focus on context rather than object names
alone. On two novel object detection benchmarks, LVIS and OminiLabel, under the
zero-shot detection setting, our approach achieves 34.8 APr minival (+9.1) and
29.3 AP (+3.6), respectively, surpassing the prior state-of-the-art models,
GLIP and FIBER, by a large margin. | cs.CV | DesCo: Learning Object Recognition with Rich Language Descriptions |
2023-06-24T20:15:07Z | http://arxiv.org/pdf/2306.14050v2 | http://arxiv.org/abs/2306.14050v2 | Liunian Harold Li, Jack Hessel, Youngjae Yu, Xiang Ren, Kai-Wei Chang, Yejin Choi | 2024-04-15T21:58:27Z | cs.CL | Chain-of-thought prompting (e.g., "Let's think step-by-step") primes large
language models to verbalize rationalization for their predictions. While
chain-of-thought can lead to dramatic performance gains, benefits appear to
emerge only for sufficiently large models (beyond 50B parameters). We show that
orders-of-magnitude smaller models (125M -- 1.3B parameters) can still benefit
from chain-of-thought prompting. To achieve this, we introduce Symbolic
Chain-of-Thought Distillation (SCoTD), a method to train a smaller student
model on rationalizations sampled from a significantly larger teacher model.
Experiments across several commonsense benchmarks show that: 1) SCoTD enhances
the performance of the student model in both supervised and few-shot settings,
and especially for challenge sets; 2) sampling many reasoning chains per
instance from the teacher is paramount; and 3) after distillation, student
chain-of-thoughts are judged by humans as comparable to the teacher, despite
orders of magnitude fewer parameters. We test several hypotheses regarding what
properties of chain-of-thought samples are important, e.g., diversity vs.
teacher likelihood vs. open-endedness. We release our corpus of
chain-of-thought samples and code. | cs.CL | Symbolic Chain-of-Thought Distillation: Small Models Can Also "Think" Step-by-Step |
2023-06-24T19:16:56Z | http://arxiv.org/pdf/2306.14040v1 | http://arxiv.org/abs/2306.14040v1 | Zeming Wei, Xiyue Zhang, Yihao Zhang, Meng Sun | 2023-06-24T19:16:56Z | cs.CL, cs.LG | Recurrent Neural Networks (RNNs) have achieved tremendous success in
processing sequential data, yet understanding and analyzing their behaviours
remains a significant challenge. To this end, many efforts have been made to
extract finite automata from RNNs, which are more amenable for analysis and
explanation. However, existing approaches like exact learning and compositional
approaches for model extraction have limitations in either scalability or
precision. In this paper, we propose a novel framework of Weighted Finite
Automata (WFA) extraction and explanation to tackle the limitations for natural
language tasks. First, to address the transition sparsity and context loss
problems we identified in WFA extraction for natural language tasks, we propose
an empirical method to complement missing rules in the transition diagram, and
adjust transition matrices to enhance the context-awareness of the WFA. We also
propose two data augmentation tactics to track more dynamic behaviours of RNN,
which further allows us to improve the extraction precision. Based on the
extracted model, we propose an explanation method for RNNs including a word
embedding method -- Transition Matrix Embeddings (TME) and TME-based task
oriented explanation for the target RNN. Our evaluation demonstrates the
advantage of our method in extraction precision than existing approaches, and
the effectiveness of TME-based explanation method in applications to
pretraining and adversarial example generation. | cs.CL | Weighted Automata Extraction and Explanation of Recurrent Neural Networks for Natural Language Tasks |
2023-06-24T18:17:38Z | http://arxiv.org/pdf/2306.14030v2 | http://arxiv.org/abs/2306.14030v2 | Tanmay Chavan, Omkar Gokhale, Aditya Kane, Shantanu Patankar, Raviraj Joshi | 2023-07-20T13:54:05Z | cs.CL, cs.LG | The research on code-mixed data is limited due to the unavailability of
dedicated code-mixed datasets and pre-trained language models. In this work, we
focus on the low-resource Indian language Marathi which lacks any prior work in
code-mixing. We present L3Cube-MeCorpus, a large code-mixed Marathi-English
(Mr-En) corpus with 10 million social media sentences for pretraining. We also
release L3Cube-MeBERT and MeRoBERTa, code-mixed BERT-based transformer models
pre-trained on MeCorpus. Furthermore, for benchmarking, we present three
supervised datasets MeHate, MeSent, and MeLID for downstream tasks like
code-mixed Mr-En hate speech detection, sentiment analysis, and language
identification respectively. These evaluation datasets individually consist of
manually annotated \url{~}12,000 Marathi-English code-mixed tweets. Ablations
show that the models trained on this novel corpus significantly outperform the
existing state-of-the-art BERT models. This is the first work that presents
artifacts for code-mixed Marathi research. All datasets and models are publicly
released at https://github.com/l3cube-pune/MarathiNLP . | cs.CL | My Boli: Code-mixed Marathi-English Corpora, Pretrained Language Models and Evaluation Benchmarks |
2023-06-24T15:27:55Z | http://arxiv.org/pdf/2306.14003v1 | http://arxiv.org/abs/2306.14003v1 | Yu Zhang, Bowen Jin, Xiusi Chen, Yanzhen Shen, Yunyi Zhang, Yu Meng, Jiawei Han | 2023-06-24T15:27:55Z | cs.CL, cs.DL, cs.LG | Instead of relying on human-annotated training samples to build a classifier,
weakly supervised scientific paper classification aims to classify papers only
using category descriptions (e.g., category names, category-indicative
keywords). Existing studies on weakly supervised paper classification are less
concerned with two challenges: (1) Papers should be classified into not only
coarse-grained research topics but also fine-grained themes, and potentially
into multiple themes, given a large and fine-grained label space; and (2) full
text should be utilized to complement the paper title and abstract for
classification. Moreover, instead of viewing the entire paper as a long linear
sequence, one should exploit the structural information such as citation links
across papers and the hierarchy of sections and paragraphs in each paper. To
tackle these challenges, in this study, we propose FUTEX, a framework that uses
the cross-paper network structure and the in-paper hierarchy structure to
classify full-text scientific papers under weak supervision. A network-aware
contrastive fine-tuning module and a hierarchy-aware aggregation module are
designed to leverage the two types of structural signals, respectively.
Experiments on two benchmark datasets demonstrate that FUTEX significantly
outperforms competitive baselines and is on par with fully supervised
classifiers that use 1,000 to 60,000 ground-truth training samples. | cs.CL | Weakly Supervised Multi-Label Classification of Full-Text Scientific Papers |
2023-06-24T14:42:43Z | http://arxiv.org/pdf/2306.13986v1 | http://arxiv.org/abs/2306.13986v1 | Alyssa Hwang, Bryan Li, Zhaoyi Hou, Dan Roth | 2023-06-24T14:42:43Z | cs.CL | With their remarkably improved text generation and prompting capabilities,
large language models can adapt existing written information into forms that
are easier to use and understand. In our work, we focus on recipes as an
example of complex, diverse, and widely used instructions. We develop a prompt
grounded in the original recipe and ingredients list that breaks recipes down
into simpler steps. We apply this prompt to recipes from various world
cuisines, and experiment with several large language models (LLMs), finding
best results with GPT-3.5. We also contribute an Amazon Mechanical Turk task
that is carefully designed to reduce fatigue while collecting human judgment of
the quality of recipe revisions. We find that annotators usually prefer the
revision over the original, demonstrating a promising application of LLMs in
serving as digital sous chefs for recipes and beyond. We release our prompt,
code, and MTurk template for public use. | cs.CL | Large Language Models as Sous Chefs: Revising Recipes with GPT-3 |
2023-06-24T13:57:32Z | http://arxiv.org/pdf/2306.13971v2 | http://arxiv.org/abs/2306.13971v2 | Xinyu Liu, Yan Ding, Kaikai An, Chunyang Xiao, Pranava Madhyastha, Tong Xiao, Jingbo Zhu | 2023-07-21T01:58:13Z | cs.CL | While state-of-the-art NLP models have demonstrated excellent performance for
aspect based sentiment analysis (ABSA), substantial evidence has been presented
on their lack of robustness. This is especially manifested as significant
degradation in performance when faced with out-of-distribution data. Recent
solutions that rely on counterfactually augmented datasets show promising
results, but they are inherently limited because of the lack of access to
explicit causal structure. In this paper, we present an alternative approach
that relies on non-counterfactual data augmentation. Our proposal instead
relies on using noisy, cost-efficient data augmentations that preserve
semantics associated with the target aspect. Our approach then relies on
modelling invariances between different versions of the data to improve
robustness. A comprehensive suite of experiments shows that our proposal
significantly improves upon strong pre-trained baselines on both standard and
robustness-specific datasets. Our approach further establishes a new
state-of-the-art on the ABSA robustness benchmark and transfers well across
domains. | cs.CL | Towards Robust Aspect-based Sentiment Analysis through Non-counterfactual Augmentations |
2023-06-24T13:51:42Z | http://arxiv.org/pdf/2306.13968v1 | http://arxiv.org/abs/2306.13968v1 | Yash Kumar Atri, Vikram Goyal, Tanmoy Chakraborty | 2023-06-24T13:51:42Z | cs.CL, cs.AI | The realm of scientific text summarization has experienced remarkable
progress due to the availability of annotated brief summaries and ample data.
However, the utilization of multiple input modalities, such as videos and
audio, has yet to be thoroughly explored. At present, scientific
multimodal-input-based text summarization systems tend to employ longer target
summaries like abstracts, leading to an underwhelming performance in the task
of text summarization.
In this paper, we deal with a novel task of extreme abstractive text
summarization (aka TL;DR generation) by leveraging multiple input modalities.
To this end, we introduce mTLDR, a first-of-its-kind dataset for the
aforementioned task, comprising videos, audio, and text, along with both
author-composed summaries and expert-annotated summaries. The mTLDR dataset
accompanies a total of 4,182 instances collected from various academic
conference proceedings, such as ICLR, ACL, and CVPR. Subsequently, we present
mTLDRgen, an encoder-decoder-based model that employs a novel dual-fused
hyper-complex Transformer combined with a Wasserstein Riemannian Encoder
Transformer, to dexterously capture the intricacies between different
modalities in a hyper-complex latent geometric space. The hyper-complex
Transformer captures the intrinsic properties between the modalities, while the
Wasserstein Riemannian Encoder Transformer captures the latent structure of the
modalities in the latent space geometry, thereby enabling the model to produce
diverse sentences. mTLDRgen outperforms 20 baselines on mTLDR as well as
another non-scientific dataset (How2) across three Rouge-based evaluation
measures. Furthermore, based on the qualitative metrics, BERTScore and FEQA,
and human evaluations, we demonstrate that the summaries generated by mTLDRgen
are fluent and congruent to the original source material. | cs.CL | Fusing Multimodal Signals on Hyper-complex Space for Extreme Abstractive Text Summarization (TL;DR) of Scientific Contents |
2023-06-24T13:22:02Z | http://arxiv.org/pdf/2306.13959v1 | http://arxiv.org/abs/2306.13959v1 | Shivani Kumar, Shubham Dudeja, Md Shad Akhtar, Tanmoy Chakraborty | 2023-06-24T13:22:02Z | cs.CL, cs.AI | In a conversational dialogue, speakers may have different emotional states
and their dynamics play an important role in understanding dialogue's emotional
discourse. However, simply detecting emotions is not sufficient to entirely
comprehend the speaker-specific changes in emotion that occur during a
conversation. To understand the emotional dynamics of speakers in an efficient
manner, it is imperative to identify the rationale or instigator behind any
changes or flips in emotion expressed by the speaker. In this paper, we explore
the task called Instigator based Emotion Flip Reasoning (EFR), which aims to
identify the instigator behind a speaker's emotion flip within a conversation.
For example, an emotion flip from joy to anger could be caused by an instigator
like threat. To facilitate this task, we present MELD-I, a dataset that
includes ground-truth EFR instigator labels, which are in line with emotional
psychology. To evaluate the dataset, we propose a novel neural architecture
called TGIF, which leverages Transformer encoders and stacked GRUs to capture
the dialogue context, speaker dynamics, and emotion sequence in a conversation.
Our evaluation demonstrates state-of-the-art performance (+4-12% increase in
F1-score) against five baselines used for the task. Further, we establish the
generalizability of TGIF on an unseen dataset in a zero-shot setting.
Additionally, we provide a detailed analysis of the competing models,
highlighting the advantages and limitations of our neural architecture. | cs.CL | Emotion Flip Reasoning in Multiparty Conversations |
2023-06-24T12:56:56Z | http://arxiv.org/pdf/2306.13954v1 | http://arxiv.org/abs/2306.13954v1 | Ridam Pal, Sanjana S, Deepak Mahto, Kriti Agrawal, Gopal Mengi, Sargun Nagpal, Akshaya Devadiga, Tavpritesh Sethi | 2023-06-24T12:56:56Z | cs.CL, cs.CY | The COVID-19 Infodemic had an unprecedented impact on health behaviors and
outcomes at a global scale. While many studies have focused on a qualitative
and quantitative understanding of misinformation, including sentiment analysis,
there is a gap in understanding the emotion-carriers of misinformation and
their differences across geographies. In this study, we characterized emotion
carriers and their impact on vaccination rates in India and the United States.
A manually labelled dataset was created from 2.3 million tweets and collated
with three publicly available datasets (CoAID, AntiVax, CMU) to train deep
learning models for misinformation classification. Misinformation labelled
tweets were further analyzed for behavioral aspects by leveraging Plutchik
Transformers to determine the emotion for each tweet. Time series analysis was
conducted to study the impact of misinformation on spatial and temporal
characteristics. Further, categorical classification was performed using
transformer models to assign categories for the misinformation tweets.
Word2Vec+BiLSTM was the best model for misinformation classification, with an
F1-score of 0.92. The US had the highest proportion of misinformation tweets
(58.02%), followed by the UK (10.38%) and India (7.33%). Disgust, anticipation,
and anger were associated with an increased prevalence of misinformation
tweets. Disgust was the predominant emotion associated with misinformation
tweets in the US, while anticipation was the predominant emotion in India. For
India, the misinformation rate exhibited a lead relationship with vaccination,
while in the US it lagged behind vaccination. Our study deciphered that
emotions acted as differential carriers of misinformation across geography and
time. These carriers can be monitored to develop strategic interventions for
countering misinformation, leading to improved public health. | cs.CL | Characterizing the Emotion Carriers of COVID-19 Misinformation and Their Impact on Vaccination Outcomes in India and the United States |
2023-06-24T12:09:43Z | http://arxiv.org/pdf/2306.13947v1 | http://arxiv.org/abs/2306.13947v1 | Muhammed Cihat Ünal, Betül Aygün, Aydın Gerek | 2023-06-24T12:09:43Z | cs.CL, cs.LG | Transformer based pre-trained models such as BERT and its variants, which are
trained on large corpora, have demonstrated tremendous success for natural
language processing (NLP) tasks. Most of academic works are based on the
English language; however, the number of multilingual and language specific
studies increase steadily. Furthermore, several studies claimed that language
specific models outperform multilingual models in various tasks. Therefore, the
community tends to train or fine-tune the models for the language of their case
study, specifically. In this paper, we focus on Turkish maps data and
thoroughly evaluate both multilingual and Turkish based BERT, DistilBERT,
ELECTRA and RoBERTa. Besides, we also propose a MultiLayer Perceptron (MLP) for
fine-tuning BERT in addition to the standard approach of one-layer fine-tuning.
For the dataset, a mid-sized Address Parsing corpus taken with a relatively
high quality is constructed. Conducted experiments on this dataset indicate
that Turkish language specific models with MLP fine-tuning yields slightly
better results when compared to the multilingual fine-tuned models. Moreover,
visualization of address tokens' representations further indicates the
effectiveness of BERT variants for classifying a variety of addresses. | cs.CL | Comparison of Pre-trained Language Models for Turkish Address Parsing |
2023-06-24T10:07:01Z | http://arxiv.org/pdf/2306.13922v1 | http://arxiv.org/abs/2306.13922v1 | Aviv Weinstein, Yoav Goldberg | 2023-06-24T10:07:01Z | cs.CL | Deverbal nouns are nominal forms of verbs commonly used in written English
texts to describe events or actions, as well as their arguments. However, many
NLP systems, and in particular pattern-based ones, neglect to handle such
nominalized constructions. The solutions that do exist for handling arguments
of nominalized constructions are based on semantic annotation and require
semantic ontologies, making their applications restricted to a small set of
nouns. We propose to adopt instead a more syntactic approach, which maps the
arguments of deverbal nouns to the universal-dependency relations of the
corresponding verbal construction. We present an unsupervised mechanism --
based on contextualized word representations -- which allows to enrich
universal-dependency trees with dependency arcs denoting arguments of deverbal
nouns, using the same labels as the corresponding verbal cases. By sharing the
same label set as in the verbal case, patterns that were developed for verbs
can be applied without modification but with high accuracy also to the nominal
constructions. | cs.CL | Unsupervised Mapping of Arguments of Deverbal Nouns to Their Corresponding Verbal Labels |
2023-06-24T08:48:24Z | http://arxiv.org/pdf/2306.13906v1 | http://arxiv.org/abs/2306.13906v1 | Jaromir Savelka, Kevin D. Ashley, Morgan A Gray, Hannes Westermann, Huihui Xu | 2023-06-24T08:48:24Z | cs.CL | We evaluated the capability of generative pre-trained transformers~(GPT-4) in
analysis of textual data in tasks that require highly specialized domain
expertise. Specifically, we focused on the task of analyzing court opinions to
interpret legal concepts. We found that GPT-4, prompted with annotation
guidelines, performs on par with well-trained law student annotators. We
observed that, with a relatively minor decrease in performance, GPT-4 can
perform batch predictions leading to significant cost reductions. However,
employing chain-of-thought prompting did not lead to noticeably improved
performance on this task. Further, we demonstrated how to analyze GPT-4's
predictions to identify and mitigate deficiencies in annotation guidelines, and
subsequently improve the performance of the model. Finally, we observed that
the model is quite brittle, as small formatting related changes in the prompt
had a high impact on the predictions. These findings can be leveraged by
researchers and practitioners who engage in semantic/pragmatic annotations of
texts in the context of the tasks requiring highly specialized domain
expertise. | cs.CL | Can GPT-4 Support Analysis of Textual Data in Tasks Requiring Highly Specialized Domain Expertise? |
2023-06-24T08:45:47Z | http://arxiv.org/pdf/2306.13905v1 | http://arxiv.org/abs/2306.13905v1 | Shreya Ghosh, Saptarshi Sengupta, Prasenjit Mitra | 2023-06-24T08:45:47Z | cs.CL, cs.AI, cs.LG | In this paper, we lay out a vision for analysing semantic trajectory traces
and generating synthetic semantic trajectory data (SSTs) using generative
language model. Leveraging the advancements in deep learning, as evident by
progress in the field of natural language processing (NLP), computer vision,
etc. we intend to create intelligent models that can study the semantic
trajectories in various contexts, predicting future trends, increasing machine
understanding of the movement of animals, humans, goods, etc. enhancing
human-computer interactions, and contributing to an array of applications
ranging from urban-planning to personalized recommendation engines and business
strategy. | cs.CL | Spatio-temporal Storytelling? Leveraging Generative Models for Semantic Trajectory Analysis |
2023-06-24T08:27:39Z | http://arxiv.org/pdf/2306.13899v1 | http://arxiv.org/abs/2306.13899v1 | Syed Rifat Raiyan, Md. Nafis Faiyaz, Shah Md. Jawad Kabir, Mohsinul Kabir, Hasan Mahmud, Md Kamrul Hasan | 2023-06-24T08:27:39Z | cs.CL, cs.AI | The art of mathematical reasoning stands as a fundamental pillar of
intellectual progress and is a central catalyst in cultivating human ingenuity.
Researchers have recently published a plethora of works centered around the
task of solving Math Word Problems (MWP) $-$ a crucial stride towards general
AI. These existing models are susceptible to dependency on shallow heuristics
and spurious correlations to derive the solution expressions. In order to
ameliorate this issue, in this paper, we propose a framework for MWP solvers
based on the generation of linguistic variants of the problem text. The
approach involves solving each of the variant problems and electing the
predicted expression with the majority of the votes. We use DeBERTa
(Decoding-enhanced BERT with disentangled attention) as the encoder to leverage
its rich textual representations and enhanced mask decoder to construct the
solution expressions. Furthermore, we introduce a challenging dataset,
$\mathrm{P\small{ARA}\normalsize{MAWPS}}$, consisting of paraphrased,
adversarial, and inverse variants of selectively sampled MWPs from the
benchmark $\mathrm{M\small{AWPS}}$ dataset. We extensively experiment on this
dataset along with other benchmark datasets using some baseline MWP solver
models. We show that training on linguistic variants of problem statements and
voting on candidate predictions improve the mathematical reasoning and
robustness of the model. We make our code and data publicly available. | cs.CL | Math Word Problem Solving by Generating Linguistic Variants of Problem Statements |
2023-06-24T07:45:38Z | http://arxiv.org/pdf/2306.13891v2 | http://arxiv.org/abs/2306.13891v2 | Yanai Elazar, Jiayao Zhang, David Wadden, Bo Zhang, Noah A. Smith | 2024-02-20T06:37:12Z | cs.CL | What is the effect of releasing a preprint of a paper before it is submitted
for peer review? No randomized controlled trial has been conducted, so we turn
to observational data to answer this question. We use data from the ICLR
conference (2018--2022) and apply methods from causal inference to estimate the
effect of arXiving a paper before the reviewing period (early arXiving) on its
acceptance to the conference. Adjusting for confounders such as topic, authors,
and quality, we may estimate the causal effect. However, since quality is a
challenging construct to estimate, we use the negative outcome control method,
using paper citation count as a control variable to debias the quality
confounding effect. Our results suggest that early arXiving may have a small
effect on a paper's chances of acceptance. However, this effect (when existing)
does not differ significantly across different groups of authors, as grouped by
author citation count and institute rank. This suggests that early arXiving
does not provide an advantage to any particular group. | cs.CL | Estimating the Causal Effect of Early ArXiving on Paper Acceptance |
2023-06-24T07:27:53Z | http://arxiv.org/pdf/2306.13888v1 | http://arxiv.org/abs/2306.13888v1 | Aabha Pingle, Aditya Vyawahare, Isha Joshi, Rahul Tangsali, Raviraj Joshi | 2023-06-24T07:27:53Z | cs.CL, cs.LG | The exploration of sentiment analysis in low-resource languages, such as
Marathi, has been limited due to the availability of suitable datasets. In this
work, we present L3Cube-MahaSent-MD, a multi-domain Marathi sentiment analysis
dataset, with four different domains - movie reviews, general tweets, TV show
subtitles, and political tweets. The dataset consists of around 60,000 manually
tagged samples covering 3 distinct sentiments - positive, negative, and
neutral. We create a sub-dataset for each domain comprising 15k samples. The
MahaSent-MD is the first comprehensive multi-domain sentiment analysis dataset
within the Indic sentiment landscape. We fine-tune different monolingual and
multilingual BERT models on these datasets and report the best accuracy with
the MahaBERT model. We also present an extensive in-domain and cross-domain
analysis thus highlighting the need for low-resource multi-domain datasets. The
data and models are available at https://github.com/l3cube-pune/MarathiNLP . | cs.CL | L3Cube-MahaSent-MD: A Multi-domain Marathi Sentiment Analysis Dataset and Transformer Models |
2023-06-24T05:02:34Z | http://arxiv.org/pdf/2306.13865v1 | http://arxiv.org/abs/2306.13865v1 | Yuxin Zi, Kaushik Roy, Vignesh Narayanan, Manas Gaur, Amit Sheth | 2023-06-24T05:02:34Z | cs.CL | Large Language Models (LLMs) encode meanings of words in the form of
distributed semantics. Distributed semantics capture common statistical
patterns among language tokens (words, phrases, and sentences) from large
amounts of data. LLMs perform exceedingly well across General Language
Understanding Evaluation (GLUE) tasks designed to test a model's understanding
of the meanings of the input tokens. However, recent studies have shown that
LLMs tend to generate unintended, inconsistent, or wrong texts as outputs when
processing inputs that were seen rarely during training, or inputs that are
associated with diverse contexts (e.g., well-known hallucination phenomenon in
language generation tasks). Crowdsourced and expert-curated knowledge graphs
such as ConceptNet are designed to capture the meaning of words from a compact
set of well-defined contexts. Thus LLMs may benefit from leveraging such
knowledge contexts to reduce inconsistencies in outputs. We propose a novel
ensemble learning method, Interpretable Ensemble Representation Learning
(IERL), that systematically combines LLM and crowdsourced knowledge
representations of input tokens. IERL has the distinct advantage of being
interpretable by design (when was the LLM context used vs. when was the
knowledge context used?) over state-of-the-art (SOTA) methods, allowing
scrutiny of the inputs in conjunction with the parameters of the model,
facilitating the analysis of models' inconsistent or irrelevant outputs.
Although IERL is agnostic to the choice of LLM and crowdsourced knowledge, we
demonstrate our approach using BERT and ConceptNet. We report improved or
competitive results with IERL across GLUE tasks over current SOTA methods and
significantly enhanced model interpretability. | cs.CL | IERL: Interpretable Ensemble Representation Learning -- Combining CrowdSourced Knowledge and Distributed Semantic Representations |
2023-06-24T02:26:45Z | http://arxiv.org/pdf/2306.13841v1 | http://arxiv.org/abs/2306.13841v1 | Brando Miranda, Patrick Yu, Saumya Goyal, Yu-Xiong Wang, Sanmi Koyejo | 2023-06-24T02:26:45Z | cs.LG, cs.AI, cs.CL, cs.CV | In the context of few-shot learning, it is currently believed that a fixed
pre-trained (PT) model, along with fine-tuning the final layer during
evaluation, outperforms standard meta-learning algorithms. We re-evaluate these
claims under an in-depth empirical examination of an extensive set of formally
diverse datasets and compare PT to Model Agnostic Meta-Learning (MAML). Unlike
previous work, we emphasize a fair comparison by using: the same architecture,
the same optimizer, and all models trained to convergence. Crucially, we use a
more rigorous statistical tool -- the effect size (Cohen's d) -- to determine
the practical significance of the difference between a model trained with PT
vs. a MAML. We then use a previously proposed metric -- the diversity
coefficient -- to compute the average formal diversity of a dataset. Using this
analysis, we demonstrate the following: 1. when the formal diversity of a data
set is low, PT beats MAML on average and 2. when the formal diversity is high,
MAML beats PT on average. The caveat is that the magnitude of the average
difference between a PT vs. MAML using the effect size is low (according to
classical statistical thresholds) -- less than 0.2. Nevertheless, this
observation is contrary to the currently held belief that a pre-trained model
is always better than a meta-learning model. Our extensive experiments consider
21 few-shot learning benchmarks, including the large-scale few-shot learning
dataset Meta-Data set. We also show no significant difference between a MAML
model vs. a PT model with GPT-2 on Openwebtext. We, therefore, conclude that a
pre-trained model does not always beat a meta-learned model and that the formal
diversity of a dataset is a driving factor. | cs.LG | Is Pre-training Truly Better Than Meta-Learning? |
2023-06-24T02:25:56Z | http://arxiv.org/pdf/2306.13840v2 | http://arxiv.org/abs/2306.13840v2 | Alycia Lee, Brando Miranda, Sudharsan Sundar, Sanmi Koyejo | 2023-09-26T23:29:05Z | cs.CL, cs.AI, cs.LG, cs.NE | Current trends to pre-train capable Large Language Models (LLMs) mostly focus
on scaling of model and dataset size. However, the quality of pre-training data
is an important factor for training powerful LLMs, yet it is a nebulous concept
that has not been fully characterized. Therefore, we use the recently proposed
Task2Vec diversity coefficient to ground and understand formal aspects of data
quality, to go beyond scale alone. Specifically, we measure the diversity
coefficient of publicly available pre-training datasets to demonstrate that
their formal diversity is high when compared to theoretical lower and upper
bounds. In addition, to build confidence in the diversity coefficient, we
conduct interpretability experiments and find that the coefficient aligns with
intuitive properties of diversity, e.g., it increases as the number of latent
concepts increases. We conclude the diversity coefficient is reliable, show
it's high for publicly available LLM datasets, and conjecture it can be used to
build useful diverse datasets for LLMs. | cs.CL | Beyond Scale: the Diversity Coefficient as a Data Quality Metric Demonstrates LLMs are Pre-trained on Formally Diverse Data |
2023-06-23T23:53:49Z | http://arxiv.org/pdf/2306.13817v1 | http://arxiv.org/abs/2306.13817v1 | Jason H. J. Lu, Qingzhen Guo | 2023-06-23T23:53:49Z | cs.AI, cs.CL, I.2 | We introduce a framework for analyzing various types of information in an NLP
Transformer. In this approach, we distinguish four layers of information:
positional, syntactic, semantic, and contextual. We also argue that the common
practice of adding positional information to semantic embedding is sub-optimal
and propose instead a Linear-and-Add approach. Our analysis reveals an
autogenetic separation of positional information through the deep layers. We
show that the distilled positional components of the embedding vectors follow
the path of a helix, both on the encoder side and on the decoder side. We
additionally show that on the encoder side, the conceptual dimensions generate
Part-of-Speech (PoS) clusters. On the decoder side, we show that a di-gram
approach helps to reveal the PoS clusters of the next token. Our approach paves
a way to elucidate the processing of information through the deep layers of an
NLP Transformer. | cs.AI | The Double Helix inside the NLP Transformer |
2023-06-23T22:38:32Z | http://arxiv.org/pdf/2306.13804v3 | http://arxiv.org/abs/2306.13804v3 | Syed Aun Muhammad Zaidi, Siddique Latif, Junaid Qadir | 2023-07-14T13:36:35Z | cs.CL, cs.SD, eess.AS | Despite the recent progress in speech emotion recognition (SER),
state-of-the-art systems are unable to achieve improved performance in
cross-language settings. In this paper, we propose a Multimodal Dual Attention
Transformer (MDAT) model to improve cross-language SER. Our model utilises
pre-trained models for multimodal feature extraction and is equipped with a
dual attention mechanism including graph attention and co-attention to capture
complex dependencies across different modalities and achieve improved
cross-language SER results using minimal target language data. In addition, our
model also exploits a transformer encoder layer for high-level feature
representation to improve emotion classification accuracy. In this way, MDAT
performs refinement of feature representation at various stages and provides
emotional salient features to the classification layer. This novel approach
also ensures the preservation of modality-specific emotional information while
enhancing cross-modality and cross-language interactions. We assess our model's
performance on four publicly available SER datasets and establish its superior
effectiveness compared to recent approaches and baseline models. | cs.CL | Cross-Language Speech Emotion Recognition Using Multimodal Dual Attention Transformers |
2023-06-23T22:10:05Z | http://arxiv.org/pdf/2306.13797v1 | http://arxiv.org/abs/2306.13797v1 | Rohitash Chandra, Jayesh Sonawane, Janhavi Lande, Cathy Yu | 2023-06-23T22:10:05Z | cs.SI, cs.CL | Anti-vaccine sentiments have been well-known and reported throughout the
history of viral outbreaks and vaccination programmes. The COVID-19 pandemic
had fear and uncertainty about vaccines which has been well expressed on social
media platforms such as Twitter. We analyse Twitter sentiments from the
beginning of the COVID-19 pandemic and study the public behaviour during the
planning, development and deployment of vaccines expressed in tweets worldwide
using a sentiment analysis framework via deep learning models. In this way, we
provide visualisation and analysis of anti-vaccine sentiments over the course
of the COVID-19 pandemic. Our results show a link between the number of tweets,
the number of cases, and the change in sentiment polarity scores during major
waves of COVID-19 cases. We also found that the first half of the pandemic had
drastic changes in the sentiment polarity scores that later stabilised which
implies that the vaccine rollout had an impact on the nature of discussions on
social media. | cs.SI | An analysis of vaccine-related sentiments from development to deployment of COVID-19 vaccines |
2023-06-23T21:25:38Z | http://arxiv.org/pdf/2306.13789v1 | http://arxiv.org/abs/2306.13789v1 | Adel Elmahdy, Ahmed Salem | 2023-06-23T21:25:38Z | cs.CL, cs.CR, cs.LG | Natural language processing (NLP) models have become increasingly popular in
real-world applications, such as text classification. However, they are
vulnerable to privacy attacks, including data reconstruction attacks that aim
to extract the data used to train the model. Most previous studies on data
reconstruction attacks have focused on LLM, while classification models were
assumed to be more secure. In this work, we propose a new targeted data
reconstruction attack called the Mix And Match attack, which takes advantage of
the fact that most classification models are based on LLM. The Mix And Match
attack uses the base model of the target model to generate candidate tokens and
then prunes them using the classification head. We extensively demonstrate the
effectiveness of the attack using both random and organic canaries. This work
highlights the importance of considering the privacy risks associated with data
reconstruction attacks in classification models and offers insights into
possible leakages. | cs.CL | Deconstructing Classifiers: Towards A Data Reconstruction Attack Against Text Classification Models |
2023-06-23T20:57:32Z | http://arxiv.org/pdf/2306.14924v1 | http://arxiv.org/abs/2306.14924v1 | Robert Chew, John Bollenbacher, Michael Wenger, Jessica Speer, Annice Kim | 2023-06-23T20:57:32Z | cs.CL, cs.AI, cs.LG, stat.AP | Deductive coding is a widely used qualitative research method for determining
the prevalence of themes across documents. While useful, deductive coding is
often burdensome and time consuming since it requires researchers to read,
interpret, and reliably categorize a large body of unstructured text documents.
Large language models (LLMs), like ChatGPT, are a class of quickly evolving AI
tools that can perform a range of natural language processing and reasoning
tasks. In this study, we explore the use of LLMs to reduce the time it takes
for deductive coding while retaining the flexibility of a traditional content
analysis. We outline the proposed approach, called LLM-assisted content
analysis (LACA), along with an in-depth case study using GPT-3.5 for LACA on a
publicly available deductive coding data set. Additionally, we conduct an
empirical benchmark using LACA on 4 publicly available data sets to assess the
broader question of how well GPT-3.5 performs across a range of deductive
coding tasks. Overall, we find that GPT-3.5 can often perform deductive coding
at levels of agreement comparable to human coders. Additionally, we demonstrate
that LACA can help refine prompts for deductive coding, identify codes for
which an LLM is randomly guessing, and help assess when to use LLMs vs. human
coders for deductive coding. We conclude with several implications for future
practice of deductive coding and related research methods. | cs.CL | LLM-Assisted Content Analysis: Using Large Language Models to Support Deductive Coding |
2023-06-23T20:14:07Z | http://arxiv.org/pdf/2306.13775v1 | http://arxiv.org/abs/2306.13775v1 | Selahattin Serdar Helli, Senem Tanberk, Sena Nur Cavsak | 2023-06-23T20:14:07Z | cs.CL, cs.CV | Information extraction (IE), one of the main tasks of natural language
processing (NLP), has recently increased importance in the use of resumes. In
studies on the text to extract information from the CV, sentence classification
was generally made using NLP models. In this study, it is aimed to extract
information by classifying all of the text groups after pre-processing such as
Optical Character Recognition (OCT) and object recognition with the YOLOv8
model of the resumes. The text dataset consists of 286 resumes collected for 5
different (education, experience, talent, personal and language) job
descriptions in the IT industry. The dataset created for object recognition
consists of 1198 resumes, which were collected from the open-source internet
and labeled as sets of text. BERT, BERT-t, DistilBERT, RoBERTa and XLNet were
used as models. F1 score variances were used to compare the model results. In
addition, the YOLOv8 model has also been reported comparatively in itself. As a
result of the comparison, DistilBERT was showed better results despite having a
lower number of parameters than other models. | cs.CL | Resume Information Extraction via Post-OCR Text Processing |
2023-06-23T18:49:20Z | http://arxiv.org/pdf/2306.13734v2 | http://arxiv.org/abs/2306.13734v2 | Samuele Cornell, Matthew Wiesner, Shinji Watanabe, Desh Raj, Xuankai Chang, Paola Garcia, Matthew Maciejewski, Yoshiki Masuyama, Zhong-Qiu Wang, Stefano Squartini, Sanjeev Khudanpur | 2023-07-14T09:45:21Z | eess.AS, cs.CL, cs.SD | The CHiME challenges have played a significant role in the development and
evaluation of robust automatic speech recognition (ASR) systems. We introduce
the CHiME-7 distant ASR (DASR) task, within the 7th CHiME challenge. This task
comprises joint ASR and diarization in far-field settings with multiple, and
possibly heterogeneous, recording devices. Different from previous challenges,
we evaluate systems on 3 diverse scenarios: CHiME-6, DiPCo, and Mixer 6. The
goal is for participants to devise a single system that can generalize across
different array geometries and use cases with no a-priori information. Another
departure from earlier CHiME iterations is that participants are allowed to use
open-source pre-trained models and datasets. In this paper, we describe the
challenge design, motivation, and fundamental research questions in detail. We
also present the baseline system, which is fully array-topology agnostic and
features multi-channel diarization, channel selection, guided source separation
and a robust ASR model that leverages self-supervised speech representations
(SSLR). | eess.AS | The CHiME-7 DASR Challenge: Distant Meeting Transcription with Multiple Devices in Diverse Scenarios |
2023-06-23T17:59:09Z | http://arxiv.org/pdf/2306.13651v2 | http://arxiv.org/abs/2306.13651v2 | Neel Jain, Khalid Saifullah, Yuxin Wen, John Kirchenbauer, Manli Shu, Aniruddha Saha, Micah Goldblum, Jonas Geiping, Tom Goldstein | 2023-06-29T17:30:45Z | cs.CL, cs.LG | With the rise of Large Language Models (LLMs) and their ubiquitous deployment
in diverse domains, measuring language model behavior on realistic data is
imperative. For example, a company deploying a client-facing chatbot must
ensure that the model will not respond to client requests with profanity.
Current evaluations approach this problem using small, domain-specific datasets
with human-curated labels. These evaluation sets are often sampled from a
narrow and simplified distribution, and data sources can unknowingly be leaked
into the training set which can lead to misleading evaluations. To bypass these
drawbacks, we propose a framework for self-supervised evaluation of LLMs by
analyzing their sensitivity or invariance to transformations on the input text.
Self-supervised evaluation can directly monitor LLM behavior on datasets
collected in the wild or streamed during live model deployment. We demonstrate
self-supervised evaluation strategies for measuring closed-book knowledge,
toxicity, and long-range context dependence, in addition to sensitivity to
grammatical structure and tokenization errors. When comparisons to similar
human-labeled benchmarks are available, we find strong correlations between
self-supervised and human-supervised evaluations. The self-supervised paradigm
complements current evaluation strategies that rely on labeled data. | cs.CL | Bring Your Own Data! Self-Supervised Evaluation for Large Language Models |
2023-06-23T17:56:26Z | http://arxiv.org/pdf/2306.13649v3 | http://arxiv.org/abs/2306.13649v3 | Rishabh Agarwal, Nino Vieillard, Yongchao Zhou, Piotr Stanczyk, Sabela Ramos, Matthieu Geist, Olivier Bachem | 2024-01-17T03:23:23Z | cs.LG, cs.AI, cs.CL | Knowledge distillation (KD) is widely used for compressing a teacher model to
reduce its inference cost and memory footprint, by training a smaller student
model. However, current KD methods for auto-regressive sequence models suffer
from distribution mismatch between output sequences seen during training and
those generated by the student during inference. To address this issue, we
introduce Generalized Knowledge Distillation (GKD). Instead of solely relying
on a fixed set of output sequences, GKD trains the student on its
self-generated output sequences by leveraging feedback from the teacher on such
sequences. Unlike supervised KD approaches, GKD also offers the flexibility to
employ alternative loss functions between the student and teacher, which can be
useful when the student lacks the expressivity to mimic the teacher's
distribution. Furthermore, GKD facilitates the seamless integration of
distillation with RL fine-tuning (RLHF). We demonstrate the efficacy of GKD for
distilling auto-regressive language models on summarization, translation, and
arithmetic reasoning tasks, and task-agnostic distillation for
instruction-tuning. | cs.LG | On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes |
2023-06-23T16:35:46Z | http://arxiv.org/pdf/2306.13596v4 | http://arxiv.org/abs/2306.13596v4 | Davoud Ataee Tarzanagh, Yingcong Li, Xuechen Zhang, Samet Oymak | 2023-12-08T18:56:20Z | cs.LG, cs.AI, cs.CL, math.OC | Attention mechanism is a central component of the transformer architecture
which led to the phenomenal success of large language models. However, the
theoretical principles underlying the attention mechanism are poorly
understood, especially its nonconvex optimization dynamics. In this work, we
explore the seminal softmax-attention model $f(\boldsymbol{X})=\langle
\boldsymbol{Xv}, \texttt{softmax}(\boldsymbol{XWp})\rangle$, where
$\boldsymbol{X}$ is the token sequence and
$(\boldsymbol{v},\boldsymbol{W},\boldsymbol{p})$ are trainable parameters. We
prove that running gradient descent on $\boldsymbol{p}$, or equivalently
$\boldsymbol{W}$, converges in direction to a max-margin solution that
separates $\textit{locally-optimal}$ tokens from non-optimal ones. This clearly
formalizes attention as an optimal token selection mechanism. Remarkably, our
results are applicable to general data and precisely characterize
$\textit{optimality}$ of tokens in terms of the value embeddings
$\boldsymbol{Xv}$ and problem geometry. We also provide a broader
regularization path analysis that establishes the margin maximizing nature of
attention even for nonlinear prediction heads. When optimizing $\boldsymbol{v}$
and $\boldsymbol{p}$ simultaneously with logistic loss, we identify conditions
under which the regularization paths directionally converge to their respective
hard-margin SVM solutions where $\boldsymbol{v}$ separates the input features
based on their labels. Interestingly, the SVM formulation of $\boldsymbol{p}$
is influenced by the support vector geometry of $\boldsymbol{v}$. Finally, we
verify our theoretical findings via numerical experiments and provide insights. | cs.LG | Max-Margin Token Selection in Attention Mechanism |
2023-06-23T16:23:24Z | http://arxiv.org/pdf/2306.15687v2 | http://arxiv.org/abs/2306.15687v2 | Matthew Le, Apoorv Vyas, Bowen Shi, Brian Karrer, Leda Sari, Rashel Moritz, Mary Williamson, Vimal Manohar, Yossi Adi, Jay Mahadeokar, Wei-Ning Hsu | 2023-10-19T13:23:28Z | eess.AS, cs.CL, cs.LG, cs.SD | Large-scale generative models such as GPT and DALL-E have revolutionized the
research community. These models not only generate high fidelity outputs, but
are also generalists which can solve tasks not explicitly taught. In contrast,
speech generative models are still primitive in terms of scale and task
generalization. In this paper, we present Voicebox, the most versatile
text-guided generative model for speech at scale. Voicebox is a
non-autoregressive flow-matching model trained to infill speech, given audio
context and text, trained on over 50K hours of speech that are not filtered or
enhanced. Similar to GPT, Voicebox can perform many different tasks through
in-context learning, but is more flexible as it can also condition on future
context. Voicebox can be used for mono or cross-lingual zero-shot
text-to-speech synthesis, noise removal, content editing, style conversion, and
diverse sample generation. In particular, Voicebox outperforms the
state-of-the-art zero-shot TTS model VALL-E on both intelligibility (5.9% vs
1.9% word error rates) and audio similarity (0.580 vs 0.681) while being up to
20 times faster. Audio samples can be found in
\url{https://voicebox.metademolab.com}. | eess.AS | Voicebox: Text-Guided Multilingual Universal Speech Generation at Scale |
2023-06-23T16:23:00Z | http://arxiv.org/pdf/2306.15686v1 | http://arxiv.org/abs/2306.15686v1 | Zhongzhi Yu, Yang Zhang, Kaizhi Qian, Yonggan Fu, Yingyan Lin | 2023-06-23T16:23:00Z | eess.AS, cs.CL | Despite the impressive performance recently achieved by automatic speech
recognition (ASR), we observe two primary challenges that hinder its broader
applications: (1) The difficulty of introducing scalability into the model to
support more languages with limited training, inference, and storage overhead;
(2) The low-resource adaptation ability that enables effective low-resource
adaptation while avoiding over-fitting and catastrophic forgetting issues.
Inspired by recent findings, we hypothesize that we can address the above
challenges with modules widely shared across languages. To this end, we propose
an ASR framework, dubbed \METHODNS, that, \textit{for the first time},
simultaneously achieves strong multilingual scalability and low-resource
adaptation ability thanks to its modularize-then-assemble strategy.
Specifically, \METHOD learns a small set of generalizable sub-modules and
adaptively assembles them for different languages to reduce the multilingual
overhead and enable effective knowledge transfer for low-resource adaptation.
Extensive experiments and visualizations demonstrate that \METHOD can
effectively discover language similarity and improve multilingual and
low-resource ASR performance over state-of-the-art (SOTA) methods, e.g., under
multilingual-ASR, our framework achieves a 0.13$\sim$2.41 lower character error
rate (CER) with 30\% smaller inference overhead over SOTA solutions on
multilingual ASR and a comparable CER, with nearly 50 times fewer trainable
parameters over SOTA solutions on low-resource tuning, respectively. | eess.AS | Master-ASR: Achieving Multilingual Scalability and Low-Resource Adaptation in ASR with Modular Learning |
2023-06-23T16:21:40Z | http://arxiv.org/pdf/2306.13588v3 | http://arxiv.org/abs/2306.13588v3 | Weizhe Yuan, Kyunghyun Cho, Jason Weston | 2024-02-03T00:24:11Z | cs.CL, cs.AI | Natural language (NL) feedback offers rich insights into user experience.
While existing studies focus on an instance-level approach, where feedback is
used to refine specific examples, we introduce a framework for system-level use
of NL feedback. We show how to use feedback to formalize system-level design
decisions in a human-in-the-loop-process -- in order to produce better models.
In particular this is done through: (i) metric design for tasks; and (ii)
language model prompt design for refining model responses. We conduct two case
studies of this approach for improving search query and dialog response
generation, demonstrating the effectiveness of system-level feedback. We show
the combination of system-level and instance-level feedback brings further
gains, and that human written instance-level feedback results in more grounded
refinements than GPT-3.5 written ones, underlying the importance of human
feedback for building systems. We release our code and data at
https://github.com/yyy-Apple/Sys-NL-Feedback. | cs.CL | System-Level Natural Language Feedback |
2023-06-23T15:21:52Z | http://arxiv.org/pdf/2306.13549v2 | http://arxiv.org/abs/2306.13549v2 | Shukang Yin, Chaoyou Fu, Sirui Zhao, Ke Li, Xing Sun, Tong Xu, Enhong Chen | 2024-04-01T17:51:54Z | cs.CV, cs.AI, cs.CL, cs.LG | Recently, Multimodal Large Language Model (MLLM) represented by GPT-4V has
been a new rising research hotspot, which uses powerful Large Language Models
(LLMs) as a brain to perform multimodal tasks. The surprising emergent
capabilities of MLLM, such as writing stories based on images and OCR-free math
reasoning, are rare in traditional multimodal methods, suggesting a potential
path to artificial general intelligence. To this end, both academia and
industry have endeavored to develop MLLMs that can compete with or even better
than GPT-4V, pushing the limit of research at a surprising speed. In this
paper, we aim to trace and summarize the recent progress of MLLMs. First of
all, we present the basic formulation of MLLM and delineate its related
concepts, including architecture, training strategy and data, as well as
evaluation. Then, we introduce research topics about how MLLMs can be extended
to support more granularity, modalities, languages, and scenarios. We continue
with multimodal hallucination and extended techniques, including Multimodal ICL
(M-ICL), Multimodal CoT (M-CoT), and LLM-Aided Visual Reasoning (LAVR). To
conclude the paper, we discuss existing challenges and point out promising
research directions. In light of the fact that the era of MLLM has only just
begun, we will keep updating this survey and hope it can inspire more research.
An associated GitHub link collecting the latest papers is available at
https://github.com/BradyFU/Awesome-Multimodal-Large-Language-Models. | cs.CV | A Survey on Multimodal Large Language Models |
2023-06-23T15:19:29Z | http://arxiv.org/pdf/2306.16282v1 | http://arxiv.org/abs/2306.16282v1 | Eli Bagno, Thierry Dana-Picard, Shulamit Reches | 2023-06-23T15:19:29Z | cs.CL | The emergence of ChatGPT has been rapid, and although it has demonstrated
positive impacts in certain domains, its influence is not universally
advantageous. Our analysis focuses on ChatGPT's capabilities in Mathematics
Education, particularly in teaching basic Linear Algebra. While there are
instances where ChatGPT delivers accurate and well-motivated answers, it is
crucial to recognize numerous cases where it makes significant mathematical
errors and fails in logical inference. These occurrences raise concerns
regarding the system's genuine understanding of mathematics, as it appears to
rely more on visual patterns rather than true comprehension. Additionally, the
suitability of ChatGPT as a teacher for students also warrants consideration. | cs.CL | ChatGPT may excel in States Medical Licensing Examination but falters in basic Linear Algebra |
2023-06-23T13:55:01Z | http://arxiv.org/pdf/2306.13501v1 | http://arxiv.org/abs/2306.13501v1 | Kaushik Roy, Yuxin Zi, Vignesh Narayanan, Manas Gaur, Amit Sheth | 2023-06-23T13:55:01Z | cs.CL | Transformer-based language models have achieved impressive success in various
natural language processing tasks due to their ability to capture complex
dependencies and contextual information using self-attention mechanisms.
However, they are not without limitations. These limitations include
hallucinations, where they produce incorrect outputs with high confidence, and
alignment issues, where they generate unhelpful and unsafe outputs for human
users. These limitations stem from the absence of implicit and missing context
in the data alone. To address this, researchers have explored augmenting these
models with external knowledge from knowledge graphs to provide the necessary
additional context. However, the ad-hoc nature of existing methods makes it
difficult to properly analyze the effects of knowledge infusion on the many
moving parts or components of a transformer. This paper introduces a systematic
method for infusing knowledge into different components of a transformer-based
model. A modular framework is proposed to identify specific components within
the transformer architecture, such as the self-attention mechanism, encoder
layers, or the input embedding layer, where knowledge infusion can be applied.
Additionally, extensive experiments are conducted on the General Language
Understanding Evaluation (GLUE) benchmark tasks, and the findings are reported.
This systematic approach aims to facilitate more principled approaches to
incorporating knowledge into language model architectures. | cs.CL | Knowledge-Infused Self Attention Transformers |
2023-06-23T12:12:08Z | http://arxiv.org/pdf/2306.13467v1 | http://arxiv.org/abs/2306.13467v1 | Pavlo Vasylenko, Pere-Lluís Huguet Cabot, Abelardo Carlos Martínez Lorenzo, Roberto Navigli | 2023-06-23T12:12:08Z | cs.CL, cs.AI | Abstract Meaning Representation (AMR) is a Semantic Parsing formalism that
aims at providing a semantic graph abstraction representing a given text.
Current approaches are based on autoregressive language models such as BART or
T5, fine-tuned through Teacher Forcing to obtain a linearized version of the
AMR graph from a sentence. In this paper, we present LeakDistill, a model and
method that explores a modification to the Transformer architecture, using
structural adapters to explicitly incorporate graph information into the
learned representations and improve AMR parsing performance. Our experiments
show how, by employing word-to-node alignment to embed graph structural
information into the encoder at training time, we can obtain state-of-the-art
AMR parsing through self-knowledge distillation, even without the use of
additional data. We release the code at
\url{http://www.github.com/sapienzanlp/LeakDistill}. | cs.CL | Incorporating Graph Information in Transformer-based AMR Parsing |
2023-06-23T12:03:07Z | http://arxiv.org/pdf/2306.13460v3 | http://arxiv.org/abs/2306.13460v3 | Zihao Yue, Anwen Hu, Liang Zhang, Qin Jin | 2023-10-28T11:44:48Z | cs.CL | Image captioning aims to describe visual content in natural language. As 'a
picture is worth a thousand words', there could be various correct descriptions
for an image. However, with maximum likelihood estimation as the training
objective, the captioning model is penalized whenever its prediction mismatches
with the label. For instance, when the model predicts a word expressing richer
semantics than the label, it will be penalized and optimized to prefer more
concise expressions, referred to as conciseness optimization. In contrast,
predictions that are more concise than labels lead to richness optimization.
Such conflicting optimization directions could eventually result in the model
generating general descriptions. In this work, we introduce Semipermeable
MaxImum Likelihood Estimation (SMILE), which allows richness optimization while
blocking conciseness optimization, thus encouraging the model to generate
longer captions with more details. Extensive experiments on two mainstream
image captioning datasets MSCOCO and Flickr30K demonstrate that SMILE
significantly enhances the descriptiveness of generated captions. We further
provide in-depth investigations to facilitate a better understanding of how
SMILE works. | cs.CL | Learning Descriptive Image Captioning via Semipermeable Maximum Likelihood Estimation |
2023-06-23T10:18:02Z | http://arxiv.org/pdf/2306.13421v1 | http://arxiv.org/abs/2306.13421v1 | Ohad Rubin, Jonathan Berant | 2023-06-23T10:18:02Z | cs.CL | Retrieval-augmented language models (LMs) have received much attention
recently. However, typically the retriever is not trained jointly as a native
component of the LM, but added to an already-pretrained LM, which limits the
ability of the LM and the retriever to adapt to one another. In this work, we
propose the Retrieval-Pretrained Transformer (RPT), an architecture and
training procedure for jointly training a retrieval-augmented LM from scratch
for the task of modeling long texts. Given a recently generated text chunk in a
long document, the LM computes query representations, which are then used to
retrieve earlier chunks in the document, located potentially tens of thousands
of tokens before. Information from retrieved chunks is fused into the LM
representations to predict the next target chunk. We train the retriever
component with a semantic objective, where the goal is to retrieve chunks that
increase the probability of the next chunk, according to a reference LM. We
evaluate RPT on four long-range language modeling tasks, spanning books, code,
and mathematical writing, and demonstrate that RPT improves retrieval quality
and subsequently perplexity across the board compared to strong baselines. | cs.CL | Long-range Language Modeling with Self-retrieval |
2023-06-23T09:30:01Z | http://arxiv.org/pdf/2306.14921v1 | http://arxiv.org/abs/2306.14921v1 | Alexander Brinkmann, Roee Shraga, Reng Chiz Der, Christian Bizer | 2023-06-23T09:30:01Z | cs.CL, cs.IR | Structured product data in the form of attribute/value pairs is the
foundation of many e-commerce applications such as faceted product search,
product comparison, and product recommendation. Product offers often only
contain textual descriptions of the product attributes in the form of titles or
free text. Hence, extracting attribute/value pairs from textual product
descriptions is an essential enabler for e-commerce applications. In order to
excel, state-of-the-art product information extraction methods require large
quantities of task-specific training data. The methods also struggle with
generalizing to out-of-distribution attributes and attribute values that were
not a part of the training data. Due to being pre-trained on huge amounts of
text as well as due to emergent effects resulting from the model size, Large
Language Models like ChatGPT have the potential to address both of these
shortcomings. This paper explores the potential of ChatGPT for extracting
attribute/value pairs from product descriptions. We experiment with different
zero-shot and few-shot prompt designs. Our results show that ChatGPT achieves a
performance similar to a pre-trained language model but requires much smaller
amounts of training data and computation for fine-tuning. | cs.CL | Product Information Extraction using ChatGPT |
2023-06-23T09:01:56Z | http://arxiv.org/pdf/2306.13379v1 | http://arxiv.org/abs/2306.13379v1 | Chieling Yueh, Evangelos Kanoulas, Bruno Martins, Camilo Thorne, Saber Akhondi | 2023-06-23T09:01:56Z | cs.CL | The high volume of published chemical patents and the importance of a timely
acquisition of their information gives rise to automating information
extraction from chemical patents. Anaphora resolution is an important component
of comprehensive information extraction, and is critical for extracting
reactions. In chemical patents, there are five anaphoric relations of interest:
co-reference, transformed, reaction associated, work up, and contained. Our
goal is to investigate how the performance of anaphora resolution models for
reaction texts in chemical patents differs in a noise-free and noisy
environment and to what extent we can improve the robustness against noise of
the model. | cs.CL | Stress Testing BERT Anaphora Resolution Models for Reaction Extraction in Chemical Patents |
2023-06-23T08:59:50Z | http://arxiv.org/pdf/2306.15685v1 | http://arxiv.org/abs/2306.15685v1 | Iuliia Nigmatulina, Srikanth Madikeri, Esaú Villatoro-Tello, Petr Motliček, Juan Zuluaga-Gomez, Karthik Pandia, Aravind Ganapathiraju | 2023-06-23T08:59:50Z | eess.AS, cs.CL | GPU decoding significantly accelerates the output of ASR predictions. While
GPUs are already being used for online ASR decoding, post-processing and
rescoring on GPUs have not been properly investigated yet. Rescoring with
available contextual information can considerably improve ASR predictions.
Previous studies have proven the viability of lattice rescoring in decoding and
biasing language model (LM) weights in offline and online CPU scenarios. In
real-time GPU decoding, partial recognition hypotheses are produced without
lattice generation, which makes the implementation of biasing more complex. The
paper proposes and describes an approach to integrate contextual biasing in
real-time GPU decoding while exploiting the standard Kaldi GPU decoder. Besides
the biasing of partial ASR predictions, our approach also permits dynamic
context switching allowing a flexible rescoring per each speech segment
directly on GPU. The code is publicly released and tested with open-sourced
test sets. | eess.AS | Implementing contextual biasing in GPU decoder for online ASR |
2023-06-23T06:33:20Z | http://arxiv.org/pdf/2306.13315v1 | http://arxiv.org/abs/2306.13315v1 | Öykü Berfin Mercan, Sena Nur Cavsak, Aysu Deliahmetoglu, Senem Tanberk | 2023-06-23T06:33:20Z | cs.CL, cs.AI | Text summarization is a fundamental task in natural language processing that
aims to condense large amounts of textual information into concise and coherent
summaries. With the exponential growth of content and the need to extract key
information efficiently, text summarization has gained significant attention in
recent years. In this study, LSTM and pre-trained T5, Pegasus, BART and
BART-Large model performances were evaluated on the open source dataset (Xsum,
CNN/Daily Mail, Amazon Fine Food Review and News Summary) and the prepared
resume dataset. This resume dataset consists of many information such as
language, education, experience, personal information, skills, and this data
includes 75 resumes. The primary objective of this research was to classify
resume text. Various techniques such as LSTM, pre-trained models, and
fine-tuned models were assessed using a dataset of resumes. The BART-Large
model fine-tuned with the resume dataset gave the best performance. | cs.CL | Abstractive Text Summarization for Resumes With Cutting Edge NLP Transformers and LSTM |
2023-06-23T06:15:54Z | http://arxiv.org/pdf/2306.13310v1 | http://arxiv.org/abs/2306.13310v1 | Chengmei Yang, Shuai Jiang, Bowei He, Chen Ma, Lianghua He | 2023-06-23T06:15:54Z | cs.CL, cs.AI | Knowledge graphs (KGs), containing many entity-relation-entity triples,
provide rich information for downstream applications. Although extracting
triples from unstructured texts has been widely explored, most of them require
a large number of labeled instances. The performance will drop dramatically
when only few labeled data are available. To tackle this problem, we propose
the Mutually Guided Few-shot learning framework for Relational Triple
Extraction (MG-FTE). Specifically, our method consists of an entity-guided
relation proto-decoder to classify the relations firstly and a relation-guided
entity proto-decoder to extract entities based on the classified relations. To
draw the connection between entity and relation, we design a proto-level fusion
module to boost the performance of both entity extraction and relation
classification. Moreover, a new cross-domain few-shot triple extraction task is
introduced. Extensive experiments show that our method outperforms many
state-of-the-art methods by 12.6 F1 score on FewRel 1.0 (single-domain) and
20.5 F1 score on FewRel 2.0 (cross-domain). | cs.CL | Mutually Guided Few-shot Learning for Relational Triple Extraction |
2023-06-23T05:55:19Z | http://arxiv.org/pdf/2306.13307v2 | http://arxiv.org/abs/2306.13307v2 | Mingyu Cui, Jiawen Kang, Jiajun Deng, Xi Yin, Yutao Xie, Xie Chen, Xunying Liu | 2023-06-26T02:48:53Z | eess.AS, cs.CL | Current ASR systems are mainly trained and evaluated at the utterance level.
Long range cross utterance context can be incorporated. A key task is to derive
a suitable compact representation of the most relevant history contexts. In
contrast to previous researches based on either LSTM-RNN encoded histories that
attenuate the information from longer range contexts, or frame level
concatenation of transformer context embeddings, in this paper compact
low-dimensional cross utterance contextual features are learned in the
Conformer-Transducer Encoder using specially designed attention pooling layers
that are applied over efficiently cached preceding utterances history vectors.
Experiments on the 1000-hr Gigaspeech corpus demonstrate that the proposed
contextualized streaming Conformer-Transducers outperform the baseline using
utterance internal context only with statistically significant WER reductions
of 0.7% to 0.5% absolute (4.3% to 3.1% relative) on the dev and test data. | eess.AS | Towards Effective and Compact Contextual Representation for Conformer Transducer Speech Recognition Systems |
2023-06-23T05:43:28Z | http://arxiv.org/pdf/2306.13304v1 | http://arxiv.org/abs/2306.13304v1 | Yuchen Zhuang, Yue Yu, Kuan Wang, Haotian Sun, Chao Zhang | 2023-06-23T05:43:28Z | cs.CL, cs.AI | Large Language Models (LLMs) have demonstrated impressive performance in
various NLP tasks, but they still suffer from challenges such as hallucination
and weak numerical reasoning. To overcome these challenges, external tools can
be used to enhance LLMs' question-answering abilities. However, current
evaluation methods do not distinguish between questions that can be answered
using LLMs' internal knowledge and those that require external information
through tool use. To address this issue, we introduce a new dataset called
ToolQA, which is designed to faithfully evaluate LLMs' ability to use external
tools for question answering. Our development of ToolQA involved a scalable,
automated process for dataset curation, along with 13 specialized tools
designed for interaction with external knowledge in order to answer questions.
Importantly, we strive to minimize the overlap between our benchmark data and
LLMs' pre-training data, enabling a more precise evaluation of LLMs' tool-use
reasoning abilities. We conducted an in-depth diagnosis of existing tool-use
LLMs to highlight their strengths, weaknesses, and potential improvements. Our
findings set a new benchmark for evaluating LLMs and suggest new directions for
future advancements. Our data and code are freely available to the broader
scientific community on GitHub. | cs.CL | ToolQA: A Dataset for LLM Question Answering with External Tools |
2023-06-22T22:29:40Z | http://arxiv.org/pdf/2306.13230v2 | http://arxiv.org/abs/2306.13230v2 | Shima Imani, Ali Beyram, Harsh Shrivastava | 2023-06-26T20:55:56Z | cs.CL, cs.AI | In this paper, we introduce DiversiGATE, a unified framework that
consolidates diverse methodologies for LLM verification. The proposed framework
comprises two main components: Diversification and Aggregation which provide a
holistic perspective on existing verification approaches, such as
Self-Consistency, Math Prompter and WebGPT. Furthermore, we propose a novel
`SelfLearner' model that conforms to the DiversiGATE framework which can learn
from its own outputs and refine its performance over time, leading to improved
accuracy. To evaluate the effectiveness of SelfLearner, we conducted a rigorous
series of experiments, including tests on synthetic data as well as on popular
arithmetic reasoning benchmarks such as GSM8K. Our results demonstrate that our
approach outperforms traditional LLMs, achieving a considerable 54.8% -> 61.8%
improvement on the GSM8K benchmark. | cs.CL | DiversiGATE: A Comprehensive Framework for Reliable Large Language Models |
2023-06-22T22:13:03Z | http://arxiv.org/pdf/2306.13213v2 | http://arxiv.org/abs/2306.13213v2 | Xiangyu Qi, Kaixuan Huang, Ashwinee Panda, Peter Henderson, Mengdi Wang, Prateek Mittal | 2023-08-16T22:38:55Z | cs.CR, cs.CL, cs.LG | Recently, there has been a surge of interest in integrating vision into Large
Language Models (LLMs), exemplified by Visual Language Models (VLMs) such as
Flamingo and GPT-4. This paper sheds light on the security and safety
implications of this trend. First, we underscore that the continuous and
high-dimensional nature of the visual input makes it a weak link against
adversarial attacks, representing an expanded attack surface of
vision-integrated LLMs. Second, we highlight that the versatility of LLMs also
presents visual attackers with a wider array of achievable adversarial
objectives, extending the implications of security failures beyond mere
misclassification. As an illustration, we present a case study in which we
exploit visual adversarial examples to circumvent the safety guardrail of
aligned LLMs with integrated vision. Intriguingly, we discover that a single
visual adversarial example can universally jailbreak an aligned LLM, compelling
it to heed a wide range of harmful instructions that it otherwise would not)
and generate harmful content that transcends the narrow scope of a `few-shot'
derogatory corpus initially employed to optimize the adversarial example. Our
study underscores the escalating adversarial risks associated with the pursuit
of multimodality. Our findings also connect the long-studied adversarial
vulnerabilities of neural networks to the nascent field of AI alignment. The
presented attack suggests a fundamental adversarial challenge for AI alignment,
especially in light of the emerging trend toward multimodality in frontier
foundation models. | cs.CR | Visual Adversarial Examples Jailbreak Aligned Large Language Models |
2023-06-22T20:38:52Z | http://arxiv.org/pdf/2306.13195v1 | http://arxiv.org/abs/2306.13195v1 | Yuetian Chen, Bowen Shi, Mei Si | 2023-06-22T20:38:52Z | cs.CL, cs.AI | Artificial intelligence has made significant progress in natural language
processing, with models like GPT-3 demonstrating impressive capabilities.
However, these models still have limitations when it comes to complex tasks
that require an understanding of the user, such as mastering human comedy
writing strategies. This paper explores humor generation using GPT-3 by
modeling human comedy writing theory and leveraging step-by-step thinking
instructions. In addition, we explore the role of cognitive distance in
creating humor. | cs.CL | Prompt to GPT-3: Step-by-Step Thinking Instructions for Humor Generation |
2023-06-22T17:47:42Z | http://arxiv.org/pdf/2306.13075v1 | http://arxiv.org/abs/2306.13075v1 | Mark Nguyen, Peter Beidler, Joseph Tsai, August Anderson, Daniel Chen, Paul Kinahan, John Kang | 2023-06-22T17:47:42Z | cs.CL, cs.CV, 68T50 (Primary), 68T10 (Secondary), I.2.7; I.5.3; J.3 | Investigators, funders, and the public desire knowledge on topics and trends
in publicly funded research but current efforts in manual categorization are
limited in scale and understanding. We developed a semi-automated approach to
extract and name research topics, and applied this to \$1.9B of NCI funding
over 21 years in the radiological sciences to determine micro- and macro-scale
research topics and funding trends. Our method relies on sequential clustering
of existing biomedical-based word embeddings, naming using subject matter
experts, and visualization to discover trends at a macroscopic scale above
individual topics. We present results using 15 and 60 cluster topics, where we
found that 2D projection of grant embeddings reveals two dominant axes:
physics-biology and therapeutic-diagnostic. For our dataset, we found that
funding for therapeutics- and physics-based research have outpaced diagnostics-
and biology-based research, respectively. We hope these results may (1) give
insight to funders on the appropriateness of their funding allocation, (2)
assist investigators in contextualizing their work and explore neighboring
research domains, and (3) allow the public to review where their tax dollars
are being allocated. | cs.CL | Semi-automated extraction of research topics and trends from NCI funding in radiological sciences from 2000-2020 |
2023-06-22T17:31:44Z | http://arxiv.org/pdf/2306.13063v2 | http://arxiv.org/abs/2306.13063v2 | Miao Xiong, Zhiyuan Hu, Xinyang Lu, Yifei Li, Jie Fu, Junxian He, Bryan Hooi | 2024-03-17T04:38:48Z | cs.CL | Empowering large language models to accurately express confidence in their
answers is essential for trustworthy decision-making. Previous confidence
elicitation methods, which primarily rely on white-box access to internal model
information or model fine-tuning, have become less suitable for LLMs,
especially closed-source commercial APIs. This leads to a growing need to
explore the untapped area of black-box approaches for LLM uncertainty
estimation. To better break down the problem, we define a systematic framework
with three components: prompting strategies for eliciting verbalized
confidence, sampling methods for generating multiple responses, and aggregation
techniques for computing consistency. We then benchmark these methods on two
key tasks-confidence calibration and failure prediction-across five types of
datasets (e.g., commonsense and arithmetic reasoning) and five widely-used LLMs
including GPT-4 and LLaMA 2 Chat. Our analysis uncovers several key insights:
1) LLMs, when verbalizing their confidence, tend to be overconfident,
potentially imitating human patterns of expressing confidence. 2) As model
capability scales up, both calibration and failure prediction performance
improve. 3) Employing our proposed strategies, such as human-inspired prompts,
consistency among multiple responses, and better aggregation strategies can
help mitigate this overconfidence from various perspectives. 4) Comparisons
with white-box methods indicate that while white-box methods perform better,
the gap is narrow, e.g., 0.522 to 0.605 in AUROC. Despite these advancements,
none of these techniques consistently outperform others, and all investigated
methods struggle in challenging tasks, such as those requiring professional
knowledge, indicating significant scope for improvement. We believe this study
can serve as a strong baseline and provide insights for eliciting confidence in
black-box LLMs. | cs.CL | Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMs |
2023-06-22T17:30:37Z | http://arxiv.org/pdf/2306.13062v1 | http://arxiv.org/abs/2306.13062v1 | Ege Kesim, Aysu Deliahmetoglu | 2023-06-22T17:30:37Z | cs.CL | Named entity recognition (NER) is used to extract information from various
documents and texts such as names and dates. It is important to extract
education and work experience information from resumes in order to filter them.
Considering the fact that all information in a resume has to be entered to the
companys system manually, automatizing this process will save time of the
companies. In this study, a deep learning-based semi-automatic named entity
recognition system has been implemented with a focus on resumes in the field of
IT. Firstly, resumes of employees from five different IT related fields has
been annotated. Six transformer based pre-trained models have been adapted to
named entity recognition problem using the annotated data. These models have
been selected among popular models in the natural language processing field.
The obtained system can recognize eight different entity types which are city,
date, degree, diploma major, job title, language, country and skill. Models
used in the experiments are compared using micro, macro and weighted F1 scores
and the performance of the methods was evaluated. Taking these scores into
account for test set the best micro and weighted F1 score is obtained by
RoBERTa and the best macro F1 score is obtained by Electra model. | cs.CL | Named entity recognition in resumes |
2023-06-22T17:13:08Z | http://arxiv.org/pdf/2306.13047v4 | http://arxiv.org/abs/2306.13047v4 | Adian Liusie, Vatsal Raina, Andrew Mullooly, Kate Knill, Mark J. F. Gales | 2023-10-15T20:58:26Z | cs.CL | Multiple choice exams are widely used to assess candidates across a diverse
range of domains and tasks. To moderate question quality, newly proposed
questions often pass through pre-test evaluation stages before being deployed
into real-world exams. Currently, this evaluation process is manually
intensive, which can lead to time lags in the question development cycle.
Streamlining this process via automation can significantly enhance efficiency,
however, there's a current lack of datasets with adequate pre-test analysis
information. In this paper we analyse a subset of the public Cambridge
Multiple-Choice Questions Reading Database released by Cambridge University
Press & Assessment; a multiple-choice comprehension dataset of questions at
different target levels, with corresponding candidate selection distributions.
We introduce the task of candidate distribution matching, propose several
evaluation metrics for the task, and demonstrate that automatic systems trained
on RACE++ can be leveraged as baselines for our task. We further demonstrate
that these automatic systems can be used for practical pre-test evaluation
tasks such as detecting underperforming distractors, where our detection
systems can automatically identify poor distractors that few candidates select. | cs.CL | Analysis of the Cambridge Multiple-Choice Questions Reading Dataset with a Focus on Candidate Response Distribution |
2023-06-22T17:07:57Z | http://arxiv.org/pdf/2306.13041v1 | http://arxiv.org/abs/2306.13041v1 | Christoph Leiter, Piyawat Lertvittayakumjorn, Marina Fomicheva, Wei Zhao, Yang Gao, Steffen Eger | 2023-06-22T17:07:57Z | cs.CL, cs.CY, cs.LG | Unlike classical lexical overlap metrics such as BLEU, most current
evaluation metrics for machine translation (for example, COMET or BERTScore)
are based on black-box large language models. They often achieve strong
correlations with human judgments, but recent research indicates that the
lower-quality classical metrics remain dominant, one of the potential reasons
being that their decision processes are more transparent. To foster more
widespread acceptance of novel high-quality metrics, explainability thus
becomes crucial. In this concept paper, we identify key properties as well as
key goals of explainable machine translation metrics and provide a
comprehensive synthesis of recent techniques, relating them to our established
goals and properties. In this context, we also discuss the latest
state-of-the-art approaches to explainable metrics based on generative models
such as ChatGPT and GPT4. Finally, we contribute a vision of next-generation
approaches, including natural language explanations. We hope that our work can
help catalyze and guide future research on explainable evaluation metrics and,
mediately, also contribute to better and more transparent machine translation
systems. | cs.CL | Towards Explainable Evaluation Metrics for Machine Translation |
2023-06-22T16:50:40Z | http://arxiv.org/pdf/2307.09630v1 | http://arxiv.org/abs/2307.09630v1 | Huda AlShuhayeb, Behrouz Minaei-Bidgoli, Mohammad E. Shenassa, Sayyed-Ali Hossayni | 2023-06-22T16:50:40Z | cs.CL | There are many complex and rich morphological subtleties in the Arabic
language, which are very useful when analyzing traditional Arabic texts,
especially in the historical and religious contexts, and help in understanding
the meaning of the texts. Vocabulary separation means separating the word into
different parts such as root and affix. In the morphological datasets, the
variety of labels and the number of data samples helps to evaluate the
morphological methods. In this paper, we present a benchmark data set for
evaluating the methods of separating Arabic words which include about 223,690
words from the book of Sharia alIslam, which have been labeled by experts. In
terms of the volume and variety of words, this dataset is superior to other
existing data sets, and as far as we know, there are no Arabic Hadith Domain
texts. To evaluate the dataset, we applied different methods such as Farasa,
Camel, Madamira, and ALP to the dataset and we reported the annotation quality
through four evaluation methods. | cs.CL | Noor-Ghateh: A Benchmark Dataset for Evaluating Arabic Word Segmenters in Hadith Domain |
2023-06-22T15:56:50Z | http://arxiv.org/pdf/2306.13000v1 | http://arxiv.org/abs/2306.13000v1 | Jonathan H. Rystrøm | 2023-06-22T15:56:50Z | cs.CY, cs.AI, cs.CL | As generative language models are deployed in ever-wider contexts, concerns
about their political values have come to the forefront with critique from all
parts of the political spectrum that the models are biased and lack neutrality.
However, the question of what neutrality is and whether it is desirable remains
underexplored. In this paper, I examine neutrality through an audit of Delphi
[arXiv:2110.07574], a large language model designed for crowdsourced ethics. I
analyse how Delphi responds to politically controversial questions compared to
different US political subgroups. I find that Delphi is poorly calibrated with
respect to confidence and exhibits a significant political skew. Based on these
results, I examine the question of neutrality from a data-feminist lens, in
terms of how notions of neutrality shift power and further marginalise unheard
voices. These findings can hopefully contribute to a more reflexive debate
about the normative questions of alignment and what role we want generative
models to play in society. | cs.CY | Apolitical Intelligence? Auditing Delphi's responses on controversial political issues in the US |
2023-06-22T15:47:36Z | http://arxiv.org/pdf/2306.12991v2 | http://arxiv.org/abs/2306.12991v2 | Yingzhi Wang, Mirco Ravanelli, Alya Yacoubi | 2023-10-20T11:31:26Z | cs.CL | Speech Emotion Recognition (SER) typically relies on utterance-level
solutions. However, emotions conveyed through speech should be considered as
discrete speech events with definite temporal boundaries, rather than
attributes of the entire utterance. To reflect the fine-grained nature of
speech emotions, we propose a new task: Speech Emotion Diarization (SED). Just
as Speaker Diarization answers the question of "Who speaks when?", Speech
Emotion Diarization answers the question of "Which emotion appears when?". To
facilitate the evaluation of the performance and establish a common benchmark
for researchers, we introduce the Zaion Emotion Dataset (ZED), an openly
accessible speech emotion dataset that includes non-acted emotions recorded in
real-life conditions, along with manually-annotated boundaries of emotion
segments within the utterance. We provide competitive baselines and open-source
the code and the pre-trained models. | cs.CL | Speech Emotion Diarization: Which Emotion Appears When? |
2023-06-22T15:40:59Z | http://arxiv.org/pdf/2306.12982v1 | http://arxiv.org/abs/2306.12982v1 | Enas Altarawneh, Ammeta Agrawal, Michael Jenkin, Manos Papagelis | 2023-06-22T15:40:59Z | cs.CL | Online conversations are particularly susceptible to derailment, which can
manifest itself in the form of toxic communication patterns like disrespectful
comments or verbal abuse. Forecasting conversation derailment predicts signs of
derailment in advance enabling proactive moderation of conversations. Current
state-of-the-art approaches to address this problem rely on sequence models
that treat dialogues as text streams. We propose a novel model based on a graph
convolutional neural network that considers dialogue user dynamics and the
influence of public perception on conversation utterances. Through empirical
evaluation, we show that our model effectively captures conversation dynamics
and outperforms the state-of-the-art models on the CGA and CMV benchmark
datasets by 1.5\% and 1.7\%, respectively. | cs.CL | Conversation Derailment Forecasting with Graph Convolutional Networks |
2023-06-22T15:10:18Z | http://arxiv.org/pdf/2306.12951v2 | http://arxiv.org/abs/2306.12951v2 | Ratanond Koonchanok, Yanling Pan, Hyeju Jang | 2024-02-13T20:53:00Z | cs.CL | ChatGPT sets a new record with the fastest-growing user base, as a chatbot
powered by a large language model (LLM). While it demonstrates state-of-the-art
capabilities in a variety of language-generation tasks, it also raises
widespread public concerns regarding its societal impact. In this paper, we
investigated public attitudes towards ChatGPT by applying natural language
processing techniques such as sentiment analysis and topic modeling to Twitter
data from December 5, 2022 to June 10, 2023. Our sentiment analysis result
indicates that the overall sentiment was largely neutral to positive, and
negative sentiments were decreasing over time. Our topic model reveals that the
most popular topics discussed were Education, Bard, Search Engines, OpenAI,
Marketing, and Cybersecurity, but the ranking varies by month. We also analyzed
the occupations of Twitter users and found that those with occupations in arts
and entertainment tweeted aboutChatGPT most frequently. Additionally, people
tended to tweet about topics relevant to their occupation. For instance,
Cybersecurity is the most discussed topic among those with occupations related
to computer and math, and Education is the most discussed topic among those in
academic and research. Overall, our exploratory study provides insights into
the public perception of ChatGPT, which could be valuable to both the general
public and developers of this technology. | cs.CL | Public Attitudes Toward ChatGPT on Twitter: Sentiments, Topics, and Occupations |
2023-06-22T14:39:04Z | http://arxiv.org/pdf/2306.12929v2 | http://arxiv.org/abs/2306.12929v2 | Yelysei Bondarenko, Markus Nagel, Tijmen Blankevoort | 2023-11-09T14:05:51Z | cs.LG, cs.AI, cs.CL, cs.CV | Transformer models have been widely adopted in various domains over the last
years, and especially large language models have advanced the field of AI
significantly. Due to their size, the capability of these networks has
increased tremendously, but this has come at the cost of a significant increase
in necessary compute. Quantization is one of the most effective ways to reduce
the computational time and memory consumption of neural networks. Many studies
have shown, however, that modern transformer models tend to learn strong
outliers in their activations, making them difficult to quantize. To retain
acceptable performance, the existence of these outliers requires activations to
be in higher bitwidth or the use of different numeric formats, extra
fine-tuning, or other workarounds. We show that strong outliers are related to
very specific behavior of attention heads that try to learn a "no-op" or just a
partial update of the residual. To achieve the exact zeros needed in the
attention matrix for a no-update, the input to the softmax is pushed to be
larger and larger during training, causing outliers in other parts of the
network. Based on these observations, we propose two simple (independent)
modifications to the attention mechanism - clipped softmax and gated attention.
We empirically show that models pre-trained using our methods learn
significantly smaller outliers while maintaining and sometimes even improving
the floating-point task performance. This enables us to quantize transformers
to full INT8 quantization of the activations without any additional effort. We
demonstrate the effectiveness of our methods on both language models (BERT,
OPT) and vision transformers. | cs.LG | Quantizable Transformers: Removing Outliers by Helping Attention Heads Do Nothing |
2023-06-22T14:37:54Z | http://arxiv.org/pdf/2306.12925v1 | http://arxiv.org/abs/2306.12925v1 | Paul K. Rubenstein, Chulayuth Asawaroengchai, Duc Dung Nguyen, Ankur Bapna, Zalán Borsos, Félix de Chaumont Quitry, Peter Chen, Dalia El Badawy, Wei Han, Eugene Kharitonov, Hannah Muckenhirn, Dirk Padfield, James Qin, Danny Rozenberg, Tara Sainath, Johan Schalkwyk, Matt Sharifi, Michelle Tadmor Ramanovich, Marco Tagliasacchi, Alexandru Tudor, Mihajlo Velimirović, Damien Vincent, Jiahui Yu, Yongqiang Wang, Vicky Zayats, Neil Zeghidour, Yu Zhang, Zhishuai Zhang, Lukas Zilka, Christian Frank | 2023-06-22T14:37:54Z | cs.CL, cs.AI, cs.SD, eess.AS, stat.ML | We introduce AudioPaLM, a large language model for speech understanding and
generation. AudioPaLM fuses text-based and speech-based language models, PaLM-2
[Anil et al., 2023] and AudioLM [Borsos et al., 2022], into a unified
multimodal architecture that can process and generate text and speech with
applications including speech recognition and speech-to-speech translation.
AudioPaLM inherits the capability to preserve paralinguistic information such
as speaker identity and intonation from AudioLM and the linguistic knowledge
present only in text large language models such as PaLM-2. We demonstrate that
initializing AudioPaLM with the weights of a text-only large language model
improves speech processing, successfully leveraging the larger quantity of text
training data used in pretraining to assist with the speech tasks. The
resulting model significantly outperforms existing systems for speech
translation tasks and has the ability to perform zero-shot speech-to-text
translation for many languages for which input/target language combinations
were not seen in training. AudioPaLM also demonstrates features of audio
language models, such as transferring a voice across languages based on a short
spoken prompt. We release examples of our method at
https://google-research.github.io/seanet/audiopalm/examples | cs.CL | AudioPaLM: A Large Language Model That Can Speak and Listen |
2023-06-22T14:31:18Z | http://arxiv.org/pdf/2306.12916v3 | http://arxiv.org/abs/2306.12916v3 | Ran Zhang, Jihed Ouni, Steffen Eger | 2024-06-02T20:38:10Z | cs.CL | While summarization has been extensively researched in natural language
processing (NLP), cross-lingual cross-temporal summarization (CLCTS) is a
largely unexplored area that has the potential to improve cross-cultural
accessibility and understanding. This paper comprehensively addresses the CLCTS
task, including dataset creation, modeling, and evaluation. We (1) build the
first CLCTS corpus with 328 instances for hDe-En (extended version with 455
instances) and 289 for hEn-De (extended version with 501 instances), leveraging
historical fiction texts and Wikipedia summaries in English and German; (2)
examine the effectiveness of popular transformer end-to-end models with
different intermediate finetuning tasks; (3) explore the potential of GPT-3.5
as a summarizer; (4) report evaluations from humans, GPT-4, and several recent
automatic evaluation metrics. Our results indicate that intermediate task
finetuned end-to-end models generate bad to moderate quality summaries while
GPT-3.5, as a zero-shot summarizer, provides moderate to good quality outputs.
GPT-3.5 also seems very adept at normalizing historical text. To assess data
contamination in GPT-3.5, we design an adversarial attack scheme in which we
find that GPT-3.5 performs slightly worse for unseen source documents compared
to seen documents. Moreover, it sometimes hallucinates when the source
sentences are inverted against its prior knowledge with a summarization
accuracy of 0.67 for plot omission, 0.71 for entity swap, and 0.53 for plot
negation. Overall, our regression results of model performances suggest that
longer, older, and more complex source texts (all of which are more
characteristic for historical language variants) are harder to summarize for
all models, indicating the difficulty of the CLCTS task. | cs.CL | Cross-lingual Cross-temporal Summarization: Dataset, Models, Evaluation |
2023-06-22T14:29:53Z | http://arxiv.org/pdf/2306.12913v1 | http://arxiv.org/abs/2306.12913v1 | Jagabandhu Mishra, Amartya Chowdhury, S. R. Mahadeva Prasanna | 2023-06-22T14:29:53Z | eess.AS, cs.CL, cs.SD | Spoken language diarization (LD) and related tasks are mostly explored using
the phonotactic approach. Phonotactic approaches mostly use explicit way of
language modeling, hence requiring intermediate phoneme modeling and
transcribed data. Alternatively, the ability of deep learning approaches to
model temporal dynamics may help for the implicit modeling of language
information through deep embedding vectors. Hence this work initially explores
the available speaker diarization frameworks that capture speaker information
implicitly to perform LD tasks. The performance of the LD system on synthetic
code-switch data using the end-to-end x-vector approach is 6.78% and 7.06%, and
for practical data is 22.50% and 60.38%, in terms of diarization error rate and
Jaccard error rate (JER), respectively. The performance degradation is due to
the data imbalance and resolved to some extent by using pre-trained wave2vec
embeddings that provide a relative improvement of 30.74% in terms of JER. | eess.AS | Implicit spoken language diarization |
2023-06-22T14:20:15Z | http://arxiv.org/pdf/2306.12907v1 | http://arxiv.org/abs/2306.12907v1 | Mingda Chen, Kevin Heffernan, Onur Çelebi, Alex Mourachko, Holger Schwenk | 2023-06-22T14:20:15Z | cs.CL | We introduce a new proxy score for evaluating bitext mining based on
similarity in a multilingual embedding space: xSIM++. In comparison to xSIM,
this improved proxy leverages rule-based approaches to extend English sentences
in any evaluation set with synthetic, hard-to-distinguish examples which more
closely mirror the scenarios we encounter during large-scale mining. We
validate this proxy by running a significant number of bitext mining
experiments for a set of low-resource languages, and subsequently train NMT
systems on the mined data. In comparison to xSIM, we show that xSIM++ is better
correlated with the downstream BLEU scores of translation systems trained on
mined bitexts, providing a reliable proxy of bitext mining performance without
needing to run expensive bitext mining pipelines. xSIM++ also reports
performance for different error types, offering more fine-grained feedback for
model development. | cs.CL | xSIM++: An Improved Proxy to Bitext Mining Performance for Low-Resource Languages |
2023-06-22T13:52:31Z | http://arxiv.org/pdf/2306.12886v2 | http://arxiv.org/abs/2306.12886v2 | Sherzod Hakimov, Gullal S. Cheema | 2024-04-07T16:01:11Z | cs.CL, cs.DL | The ongoing Russo-Ukrainian conflict has been a subject of intense media
coverage worldwide. Understanding the global narrative surrounding this topic
is crucial for researchers that aim to gain insights into its multifaceted
dimensions. In this paper, we present a novel multimedia dataset that focuses
on this topic by collecting and processing tweets posted by news or media
companies on social media across the globe. We collected tweets from February
2022 to May 2023 to acquire approximately 1.5 million tweets in 60 different
languages along with their images. Each entry in the dataset is accompanied by
processed tags, allowing for the identification of entities, stances, textual
or visual concepts, and sentiment. The availability of this multimedia dataset
serves as a valuable resource for researchers aiming to investigate the global
narrative surrounding the ongoing conflict from various aspects such as who are
the prominent entities involved, what stances are taken, where do these stances
originate from, how are the different textual and visual concepts related to
the event portrayed. | cs.CL | Unveiling Global Narratives: A Multilingual Twitter Dataset of News Media on the Russo-Ukrainian Conflict |
2023-06-22T13:21:20Z | http://arxiv.org/pdf/2307.01202v1 | http://arxiv.org/abs/2307.01202v1 | Stephen Yang | 2023-06-22T13:21:20Z | cs.LG, cs.AI, cs.CL, cs.DL | Analysis of innovation has been fundamentally limited by conventional
approaches to broad, structural variables. This paper pushes the boundaries,
taking an LLM approach to patent analysis with the groundbreaking ChatGPT
technology. OpenAI's state-of-the-art textual embedding accesses complex
information about the quality and impact of each invention to power deep
learning predictive models. The nuanced embedding drives a 24% incremental
improvement in R-squared predicting patent value and clearly isolates the worst
and best applications. These models enable a revision of the contemporary
Kogan, Papanikolaou, Seru, and Stoffman (2017) valuation of patents by a median
deviation of 1.5 times, accounting for potential institutional predictions.
Furthermore, the market fails to incorporate timely information about
applications; a long-short portfolio based on predicted acceptance rates
achieves significant abnormal returns of 3.3% annually. The models provide an
opportunity to revolutionize startup and small-firm corporate policy vis-a-vis
patenting. | cs.LG | Predictive Patentomics: Forecasting Innovation Success and Valuation with ChatGPT |
2023-06-22T12:10:41Z | http://arxiv.org/pdf/2306.12834v1 | http://arxiv.org/abs/2306.12834v1 | Elias Hossain, Rajib Rana, Niall Higgins, Jeffrey Soar, Prabal Datta Barua, Anthony R. Pisani, Ph. D, Kathryn Turner} | 2023-06-22T12:10:41Z | cs.CL, cs.CY | Background: Natural Language Processing (NLP) is widely used to extract
clinical insights from Electronic Health Records (EHRs). However, the lack of
annotated data, automated tools, and other challenges hinder the full
utilisation of NLP for EHRs. Various Machine Learning (ML), Deep Learning (DL)
and NLP techniques are studied and compared to understand the limitations and
opportunities in this space comprehensively.
Methodology: After screening 261 articles from 11 databases, we included 127
papers for full-text review covering seven categories of articles: 1) medical
note classification, 2) clinical entity recognition, 3) text summarisation, 4)
deep learning (DL) and transfer learning architecture, 5) information
extraction, 6) Medical language translation and 7) other NLP applications. This
study follows the Preferred Reporting Items for Systematic Reviews and
Meta-Analyses (PRISMA) guidelines.
Result and Discussion: EHR was the most commonly used data type among the
selected articles, and the datasets were primarily unstructured. Various ML and
DL methods were used, with prediction or classification being the most common
application of ML or DL. The most common use cases were: the International
Classification of Diseases, Ninth Revision (ICD-9) classification, clinical
note analysis, and named entity recognition (NER) for clinical descriptions and
research on psychiatric disorders.
Conclusion: We find that the adopted ML models were not adequately assessed.
In addition, the data imbalance problem is quite important, yet we must find
techniques to address this underlining problem. Future studies should address
key limitations in studies, primarily identifying Lupus Nephritis, Suicide
Attempts, perinatal self-harmed and ICD-9 classification. | cs.CL | Natural Language Processing in Electronic Health Records in Relation to Healthcare Decision-making: A Systematic Review |
2023-06-22T10:50:23Z | http://arxiv.org/pdf/2306.12794v3 | http://arxiv.org/abs/2306.12794v3 | Mario Rodríguez-Cantelar, Chen Zhang, Chengguang Tang, Ke Shi, Sarik Ghazarian, João Sedoc, Luis Fernando D'Haro, Alexander Rudnicky | 2023-09-14T01:33:36Z | cs.CL | The advent and fast development of neural networks have revolutionized the
research on dialogue systems and subsequently have triggered various challenges
regarding their automatic evaluation. Automatic evaluation of open-domain
dialogue systems as an open challenge has been the center of the attention of
many researchers. Despite the consistent efforts to improve automatic metrics'
correlations with human evaluation, there have been very few attempts to assess
their robustness over multiple domains and dimensions. Also, their focus is
mainly on the English language. All of these challenges prompt the development
of automatic evaluation metrics that are reliable in various domains,
dimensions, and languages. This track in the 11th Dialogue System Technology
Challenge (DSTC11) is part of the ongoing effort to promote robust and
multilingual automatic evaluation metrics. This article describes the datasets
and baselines provided to participants and discusses the submission and result
details of the two proposed subtasks. | cs.CL | Overview of Robust and Multilingual Automatic Evaluation Metrics for Open-Domain Dialogue Systems at DSTC 11 Track 4 |
2023-06-22T09:42:54Z | http://arxiv.org/pdf/2306.12766v1 | http://arxiv.org/abs/2306.12766v1 | Julien Romero, Simon Razniewski | 2023-06-22T09:42:54Z | cs.CL | Structured knowledge bases (KBs) are the backbone of many
know\-ledge-intensive applications, and their automated construction has
received considerable attention. In particular, open information extraction
(OpenIE) is often used to induce structure from a text. However, although it
allows high recall, the extracted knowledge tends to inherit noise from the
sources and the OpenIE algorithm. Besides, OpenIE tuples contain an open-ended,
non-canonicalized set of relations, making the extracted knowledge's downstream
exploitation harder. In this paper, we study the problem of mapping an open KB
into the fixed schema of an existing KB, specifically for the case of
commonsense knowledge. We propose approaching the problem by generative
translation, i.e., by training a language model to generate fixed-schema
assertions from open ones. Experiments show that this approach occupies a sweet
spot between traditional manual, rule-based, or classification-based
canonicalization and purely generative KB construction like COMET. Moreover, it
produces higher mapping accuracy than the former while avoiding the
association-based noise of the latter. | cs.CL | Mapping and Cleaning Open Commonsense Knowledge Bases with Generative Translation |
2023-06-22T09:18:52Z | http://arxiv.org/pdf/2306.12756v1 | http://arxiv.org/abs/2306.12756v1 | Yu-An Liu, Ruqing Zhang, Jiafeng Guo, Wei Chen, Xueqi Cheng | 2023-06-22T09:18:52Z | cs.IR, cs.AI, cs.CL, cs.LG | Recently, we have witnessed generative retrieval increasingly gaining
attention in the information retrieval (IR) field, which retrieves documents by
directly generating their identifiers. So far, much effort has been devoted to
developing effective generative retrieval models. There has been less attention
paid to the robustness perspective. When a new retrieval paradigm enters into
the real-world application, it is also critical to measure the
out-of-distribution (OOD) generalization, i.e., how would generative retrieval
models generalize to new distributions. To answer this question, firstly, we
define OOD robustness from three perspectives in retrieval problems: 1) The
query variations; 2) The unforeseen query types; and 3) The unforeseen tasks.
Based on this taxonomy, we conduct empirical studies to analyze the OOD
robustness of several representative generative retrieval models against dense
retrieval models. The empirical results indicate that the OOD robustness of
generative retrieval models requires enhancement. We hope studying the OOD
robustness of generative retrieval models would be advantageous to the IR
community. | cs.IR | On the Robustness of Generative Retrieval Models: An Out-of-Distribution Perspective |
2023-06-22T07:57:19Z | http://arxiv.org/pdf/2306.12725v4 | http://arxiv.org/abs/2306.12725v4 | Senbao Shi, Zhenran Xu, Baotian Hu, Min Zhang | 2024-03-20T01:30:41Z | cs.CL | Multimodal Entity Linking (MEL) is the task of mapping mentions with
multimodal contexts to the referent entities from a knowledge base. Existing
MEL methods mainly focus on designing complex multimodal interaction mechanisms
and require fine-tuning all model parameters, which can be prohibitively costly
and difficult to scale in the era of Large Language Models (LLMs). In this
work, we propose GEMEL, a Generative Multimodal Entity Linking framework based
on LLMs, which directly generates target entity names. We keep the vision and
language model frozen and only train a feature mapper to enable cross-modality
interactions. To adapt LLMs to the MEL task, we leverage the in-context
learning capability of LLMs by retrieving multimodal instances as
demonstrations. Extensive experiments show that, with only ~0.3% of the model
parameters fine-tuned, GEMEL achieves state-of-the-art results on two
well-established MEL datasets (7.7% accuracy gains on WikiDiverse and 8.8%
accuracy gains on WikiMEL). The performance gain stems from mitigating the
popularity bias of LLM predictions and disambiguating less common entities
effectively. Further analysis verifies the generality and scalability of GEMEL.
Our framework is compatible with any off-the-shelf language model, paving the
way towards an efficient and general solution for utilizing LLMs in the MEL
task. Our code is available at https://github.com/HITsz-TMG/GEMEL. | cs.CL | Generative Multimodal Entity Linking |
2023-06-22T07:52:34Z | http://arxiv.org/pdf/2306.12719v1 | http://arxiv.org/abs/2306.12719v1 | Soichiro Murakami, Sho Hoshino, Peinan Zhang | 2023-06-22T07:52:34Z | cs.CL | Natural language generation methods have emerged as effective tools to help
advertisers increase the number of online advertisements they produce. This
survey entails a review of the research trends on this topic over the past
decade, from template-based to extractive and abstractive approaches using
neural networks. Additionally, key challenges and directions revealed through
the survey, including metric optimization, faithfulness, diversity,
multimodality, and the development of benchmark datasets, are discussed. | cs.CL | Natural Language Generation for Advertising: A Survey |
2023-06-22T06:44:09Z | http://arxiv.org/pdf/2306.12693v1 | http://arxiv.org/abs/2306.12693v1 | Sudhansu Bala Das, Divyajyoti Panda, Tapas Kumar Mishra, Bidyut Kr. Patra, Asif Ekbal | 2023-06-22T06:44:09Z | cs.CL | This paper gives an Indic-to-Indic (IL-IL) MNMT baseline model for 11 ILs
implemented on the Samanantar corpus and analyzed on the Flores-200 corpus. All
the models are evaluated using the BLEU score. In addition, the languages are
classified under three groups namely East Indo- Aryan (EI), Dravidian (DR), and
West Indo-Aryan (WI). The effect of language relatedness on MNMT model
efficiency is studied. Owing to the presence of large corpora from English (EN)
to ILs, MNMT IL-IL models using EN as a pivot are also built and examined. To
achieve this, English- Indic (EN-IL) models are also developed, with and
without the usage of related languages. Results reveal that using related
languages is beneficial for the WI group only, while it is detrimental for the
EI group and shows an inconclusive effect on the DR group, but it is useful for
EN-IL models. Thus, related language groups are used to develop pivot MNMT
models. Furthermore, the IL corpora are transliterated from the corresponding
scripts to a modified ITRANS script, and the best MNMT models from the previous
approaches are built on the transliterated corpus. It is observed that the
usage of pivot models greatly improves MNMT baselines with AS-TA achieving the
minimum BLEU score and PA-HI achieving the maximum score. Among languages, AS,
ML, and TA achieve the lowest BLEU score, whereas HI, PA, and GU perform the
best. Transliteration also helps the models with few exceptions. The best
increment of scores is observed in ML, TA, and BN and the worst average
increment is observed in KN, HI, and PA, across all languages. The best model
obtained is the PA-HI language pair trained on PAWI transliterated corpus which
gives 24.29 BLEU. | cs.CL | Multilingual Neural Machine Translation System for Indic to Indic Languages |
2023-06-22T06:23:31Z | http://arxiv.org/pdf/2306.12689v1 | http://arxiv.org/abs/2306.12689v1 | Andrew Kean Gao | 2023-06-22T06:23:31Z | cs.CL, cs.AI, cs.IR, cs.LG, I.2.7; D.2.12 | Vector embeddings have become ubiquitous tools for many language-related
tasks. A leading embedding model is OpenAI's text-ada-002 which can embed
approximately 6,000 words into a 1,536-dimensional vector. While powerful,
text-ada-002 is not open source and is only available via API. We trained a
simple neural network to convert open-source 768-dimensional MPNet embeddings
into text-ada-002 embeddings. We compiled a subset of 50,000 online food
reviews. We calculated MPNet and text-ada-002 embeddings for each review and
trained a simple neural network to for 75 epochs. The neural network was
designed to predict the corresponding text-ada-002 embedding for a given MPNET
embedding. Our model achieved an average cosine similarity of 0.932 on 10,000
unseen reviews in our held-out test dataset. We manually assessed the quality
of our predicted embeddings for vector search over text-ada-002-embedded
reviews. While not as good as real text-ada-002 embeddings, predicted
embeddings were able to retrieve highly relevant reviews. Our final model,
Vec2Vec, is lightweight (<80 MB) and fast. Future steps include training a
neural network with a more sophisticated architecture and a larger dataset of
paired embeddings to achieve greater performance. The ability to convert
between and align embedding spaces may be helpful for interoperability,
limiting dependence on proprietary models, protecting data privacy, reducing
costs, and offline operations. | cs.CL | Vec2Vec: A Compact Neural Network Approach for Transforming Text Embeddings with High Fidelity |
2023-06-22T05:51:22Z | http://arxiv.org/pdf/2306.12679v2 | http://arxiv.org/abs/2306.12679v2 | Mojtaba Mazoochi, Leila Rabiei, Farzaneh Rahmani, Zeinab Rajabi | 2024-03-07T04:25:50Z | cs.CL | Introduction: Microblogging websites have massed rich data sources for
sentiment analysis and opinion mining. In this regard, sentiment classification
has frequently proven inefficient because microblog posts typically lack
syntactically consistent terms and representatives since users on these social
networks do not like to write lengthy statements. Also, there are some
limitations to low-resource languages. The Persian language has exceptional
characteristics and demands unique annotated data and models for the sentiment
analysis task, which are distinctive from text features within the English
dialect. Method: This paper first constructs a user opinion dataset called
ITRC-Opinion in a collaborative environment and insource way. Our dataset
contains 60,000 informal and colloquial Persian texts from social microblogs
such as Twitter and Instagram. Second, this study proposes a new architecture
based on the convolutional neural network (CNN) model for more effective
sentiment analysis of colloquial text in social microblog posts. The
constructed datasets are used to evaluate the presented architecture.
Furthermore, some models, such as LSTM, CNN-RNN, BiLSTM, and BiGRU with
different word embeddings, including Fasttext, Glove, and Word2vec,
investigated our dataset and evaluated the results. Results: The results
demonstrate the benefit of our dataset and the proposed model (72% accuracy),
displaying meaningful improvement in sentiment classification performance. | cs.CL | Constructing Colloquial Dataset for Persian Sentiment Analysis of Social Microblogs |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.