Titles
stringlengths 6
220
| Abstracts
stringlengths 37
3.26k
| Years
int64 1.99k
2.02k
| Categories
stringclasses 1
value |
|---|---|---|---|
Approximation-Aware Dependency Parsing by Belief Propagation | We show how to train the fast dependency parser of Smith and Eisner (2008)
for improved accuracy. This parser can consider higher-order interactions among
edges while retaining O(n^3) runtime. It outputs the parse with maximum
expected recall -- but for speed, this expectation is taken under a posterior
distribution that is constructed only approximately, using loopy belief
propagation through structured factors. We show how to adjust the model
parameters to compensate for the errors introduced by this approximation, by
following the gradient of the actual loss on training data. We find this
gradient by back-propagation. That is, we treat the entire parser
(approximations and all) as a differentiable circuit, as Stoyanov et al. (2011)
and Domke (2010) did for loopy CRFs. The resulting trained parser obtains
higher accuracy with fewer iterations of belief propagation than one trained by
conditional log-likelihood.
| 2,015 | Computation and Language |
Removing Biases from Trainable MT Metrics by Using Self-Training | Most trainable machine translation (MT) metrics train their weights on human
judgments of state-of-the-art MT systems outputs. This makes trainable metrics
biases in many ways. One of them is preferring longer translations. These
biased metrics when used for tuning are evaluating different types of
translations -- n-best lists of translations with very diverse quality. Systems
tuned with these metrics tend to produce overly long translations that are
preferred by the metric but not by humans. This is usually solved by manually
tweaking metric's weights to equally value recall and precision. Our solution
is more general: (1) it does not address only the recall bias but also all
other biases that might be present in the data and (2) it does not require any
knowledge of the types of features used which is useful in cases when manual
tuning of metric's weights is not possible. This is accomplished by
self-training on unlabeled n-best lists by using metric that was initially
trained on standard human judgments. One way of looking at this is as domain
adaptation from the domain of state-of-the-art MT translations to diverse
n-best list translations.
| 2,015 | Computation and Language |
Syntax Evolution: Problems and Recursion | To investigate the evolution of syntax, we need to ascertain the evolutionary
r\^ole of syntax and, before that, the very nature of syntax. Here, we will
assume that syntax is computing. And then, since we are computationally Turing
complete, we meet an evolutionary anomaly, the anomaly of sytax: we are
syntactically too competent for syntax. Assuming that problem solving is
computing, and realizing that the evolutionary advantage of Turing completeness
is full problem solving and not syntactic proficiency, we explain the anomaly
of syntax by postulating that syntax and problem solving co-evolved in humans
towards Turing completeness. Examining the requirements that full problem
solving impose on language, we find firstly that semantics is not sufficient
and that syntax is necessary to represent problems. Our final conclusion is
that full problem solving requires a functional semantics on an infinite
tree-structured syntax. Besides these results, the introduction of Turing
completeness and problem solving to explain the evolution of syntax should help
us to fit the evolution of language within the evolution of cognition, giving
us some new clues to understand the elusive relation between language and
thinking.
| 2,019 | Computation and Language |
Information-theoretical analysis of the statistical dependencies among
three variables: Applications to written language | We develop the information-theoretical concepts required to study the
statistical dependencies among three variables. Some of such dependencies are
pure triple interactions, in the sense that they cannot be explained in terms
of a combination of pairwise correlations. We derive bounds for triple
dependencies, and characterize the shape of the joint probability distribution
of three binary variables with high triple interaction. The analysis also
allows us to quantify the amount of redundancy in the mutual information
between pairs of variables, and to assess whether the information between two
variables is or is not mediated by a third variable. These concepts are applied
to the analysis of written texts. We find that the probability that a given
word is found in a particular location within the text is not only modulated by
the presence or absence of other nearby words, but also, on the presence or
absence of nearby pairs of words. We identify the words enclosing the key
semantic concepts of the text, the triplets of words with high pairwise and
triple interactions, and the words that mediate the pairwise interactions
between other words.
| 2,015 | Computation and Language |
Classifying Relations via Long Short Term Memory Networks along Shortest
Dependency Path | Relation classification is an important research arena in the field of
natural language processing (NLP). In this paper, we present SDP-LSTM, a novel
neural network to classify the relation of two entities in a sentence. Our
neural architecture leverages the shortest dependency path (SDP) between two
entities; multichannel recurrent neural networks, with long short term memory
(LSTM) units, pick up heterogeneous information along the SDP. Our proposed
model has several distinct features: (1) The shortest dependency paths retain
most relevant information (to relation classification), while eliminating
irrelevant words in the sentence. (2) The multichannel LSTM networks allow
effective information integration from heterogeneous sources over the
dependency paths. (3) A customized dropout strategy regularizes the neural
network to alleviate overfitting. We test our model on the SemEval 2010
relation classification task, and achieve an $F_1$-score of 83.7\%, higher than
competing methods in the literature.
| 2,015 | Computation and Language |
A Comparative Study on Regularization Strategies for Embedding-based
Neural Networks | This paper aims to compare different regularization strategies to address a
common phenomenon, severe overfitting, in embedding-based neural networks for
NLP. We chose two widely studied neural models and tasks as our testbed. We
tried several frequently applied or newly proposed regularization strategies,
including penalizing weights (embeddings excluded), penalizing embeddings,
re-embedding words, and dropout. We also emphasized on incremental
hyperparameter tuning, and combining different regularizations. The results
provide a picture on tuning hyperparameters for neural NLP models.
| 2,015 | Computation and Language |
A Generative Word Embedding Model and its Low Rank Positive Semidefinite
Solution | Most existing word embedding methods can be categorized into Neural Embedding
Models and Matrix Factorization (MF)-based methods. However some models are
opaque to probabilistic interpretation, and MF-based methods, typically solved
using Singular Value Decomposition (SVD), may incur loss of corpus information.
In addition, it is desirable to incorporate global latent factors, such as
topics, sentiments or writing styles, into the word embedding model. Since
generative models provide a principled way to incorporate latent factors, we
propose a generative word embedding model, which is easy to interpret, and can
serve as a basis of more sophisticated latent factor models. The model
inference reduces to a low rank weighted positive semidefinite approximation
problem. Its optimization is approached by eigendecomposition on a submatrix,
followed by online blockwise regression, which is scalable and avoids the
information loss in SVD. In experiments on 7 common benchmark datasets, our
vectors are competitive to word2vec, and better than other MF-based methods.
| 2,015 | Computation and Language |
Online Representation Learning in Recurrent Neural Language Models | We investigate an extension of continuous online learning in recurrent neural
network language models. The model keeps a separate vector representation of
the current unit of text being processed and adaptively adjusts it after each
prediction. The initial experiments give promising results, indicating that the
method is able to increase language modelling accuracy, while also decreasing
the parameters needed to store the model along with the computation required at
each step.
| 2,017 | Computation and Language |
Effective Approaches to Attention-based Neural Machine Translation | An attentional mechanism has lately been used to improve neural machine
translation (NMT) by selectively focusing on parts of the source sentence
during translation. However, there has been little work exploring useful
architectures for attention-based NMT. This paper examines two simple and
effective classes of attentional mechanism: a global approach which always
attends to all source words and a local one that only looks at a subset of
source words at a time. We demonstrate the effectiveness of both approaches
over the WMT translation tasks between English and German in both directions.
With local attention, we achieve a significant gain of 5.0 BLEU points over
non-attentional systems which already incorporate known techniques such as
dropout. Our ensemble model using different attention architectures has
established a new state-of-the-art result in the WMT'15 English to German
translation task with 25.9 BLEU points, an improvement of 1.0 BLEU points over
the existing best system backed by NMT and an n-gram reranker.
| 2,015 | Computation and Language |
Molding CNNs for text: non-linear, non-consecutive convolutions | The success of deep learning often derives from well-chosen operational
building blocks. In this work, we revise the temporal convolution operation in
CNNs to better adapt it to text processing. Instead of concatenating word
representations, we appeal to tensor algebra and use low-rank n-gram tensors to
directly exploit interactions between words already at the convolution stage.
Moreover, we extend the n-gram convolution to non-consecutive words to
recognize patterns with intervening words. Through a combination of low-rank
tensors, and pattern weighting, we can efficiently evaluate the resulting
convolution operation via dynamic programming. We test the resulting
architecture on standard sentiment classification and news categorization
tasks. Our model achieves state-of-the-art performance both in terms of
accuracy and training speed. For instance, we obtain 51.2% accuracy on the
fine-grained sentiment classification task.
| 2,015 | Computation and Language |
Learning Meta-Embeddings by Using Ensembles of Embedding Sets | Word embeddings -- distributed representations of words -- in deep learning
are beneficial for many tasks in natural language processing (NLP). However,
different embedding sets vary greatly in quality and characteristics of the
captured semantics. Instead of relying on a more advanced algorithm for
embedding learning, this paper proposes an ensemble approach of combining
different public embedding sets with the aim of learning meta-embeddings.
Experiments on word similarity and analogy tasks and on part-of-speech tagging
show better performance of meta-embeddings compared to individual embedding
sets. One advantage of meta-embeddings is the increased vocabulary coverage. We
will release our meta-embeddings publicly.
| 2,015 | Computation and Language |
Probabilistic Modelling of Morphologically Rich Languages | This thesis investigates how the sub-structure of words can be accounted for
in probabilistic models of language. Such models play an important role in
natural language processing tasks such as translation or speech recognition,
but often rely on the simplistic assumption that words are opaque symbols. This
assumption does not fit morphologically complex language well, where words can
have rich internal structure and sub-word elements are shared across distinct
word forms.
Our approach is to encode basic notions of morphology into the assumptions of
three different types of language models, with the intention that leveraging
shared sub-word structure can improve model performance and help overcome data
sparsity that arises from morphological processes.
In the context of n-gram language modelling, we formulate a new Bayesian
model that relies on the decomposition of compound words to attain better
smoothing, and we develop a new distributed language model that learns vector
representations of morphemes and leverages them to link together
morphologically related words. In both cases, we show that accounting for word
sub-structure improves the models' intrinsic performance and provides benefits
when applied to other tasks, including machine translation.
We then shift the focus beyond the modelling of word sequences and consider
models that automatically learn what the sub-word elements of a given language
are, given an unannotated list of words. We formulate a novel model that can
learn discontiguous morphemes in addition to the more conventional contiguous
morphemes that most previous models are limited to. This approach is
demonstrated on Semitic languages, and we find that modelling discontiguous
sub-word structures leads to improvements in the task of segmenting words into
their contiguous morphemes.
| 2,015 | Computation and Language |
End-to-End Attention-based Large Vocabulary Speech Recognition | Many of the current state-of-the-art Large Vocabulary Continuous Speech
Recognition Systems (LVCSR) are hybrids of neural networks and Hidden Markov
Models (HMMs). Most of these systems contain separate components that deal with
the acoustic modelling, language modelling and sequence decoding. We
investigate a more direct approach in which the HMM is replaced with a
Recurrent Neural Network (RNN) that performs sequence prediction directly at
the character level. Alignment between the input features and the desired
character sequence is learned automatically by an attention mechanism built
into the RNN. For each predicted character, the attention mechanism scans the
input sequence and chooses relevant frames. We propose two methods to speed up
this operation: limiting the scan to a subset of most promising frames and
pooling over time the information contained in neighboring frames, thereby
reducing source sequence length. Integrating an n-gram language model into the
decoding process yields recognition accuracies similar to other HMM-free
RNN-based approaches.
| 2,016 | Computation and Language |
Exploring Metaphorical Senses and Word Representations for Identifying
Metonyms | A metonym is a word with a figurative meaning, similar to a metaphor. Because
metonyms are closely related to metaphors, we apply features that are used
successfully for metaphor recognition to the task of detecting metonyms. On the
ACL SemEval 2007 Task 8 data with gold standard metonym annotations, our system
achieved 86.45% accuracy on the location metonyms. Our code can be found on
GitHub.
| 2,015 | Computation and Language |
Recognizing Extended Spatiotemporal Expressions by Actively Trained
Average Perceptron Ensembles | Precise geocoding and time normalization for text requires that location and
time phrases be identified. Many state-of-the-art geoparsers and temporal
parsers suffer from low recall. Categories commonly missed by parsers are:
nouns used in a non- spatiotemporal sense, adjectival and adverbial phrases,
prepositional phrases, and numerical phrases. We collected and annotated data
set by querying commercial web searches API with such spatiotemporal
expressions as were missed by state-of-the- art parsers. Due to the high cost
of sentence annotation, active learning was used to label training data, and a
new strategy was designed to better select training examples to reduce labeling
cost. For the learning algorithm, we applied an average perceptron trained
Featurized Hidden Markov Model (FHMM). Five FHMM instances were used to create
an ensemble, with the output phrase selected by voting. Our ensemble model was
tested on a range of sequential labeling tasks, and has shown competitive
performance. Our contributions include (1) an new dataset annotated with named
entities and expanded spatiotemporal expressions; (2) a comparison of inference
algorithms for ensemble models showing the superior accuracy of Belief
Propagation over Viterbi Decoding; (3) a new example re-weighting method for
active ensemble learning that 'memorizes' the latest examples trained; (4) a
spatiotemporal parser that jointly recognizes expanded spatiotemporal
expressions as well as named entities.
| 2,015 | Computation and Language |
Fast, Flexible Models for Discovering Topic Correlation across
Weakly-Related Collections | Weak topic correlation across document collections with different numbers of
topics in individual collections presents challenges for existing
cross-collection topic models. This paper introduces two probabilistic topic
models, Correlated LDA (C-LDA) and Correlated HDP (C-HDP). These address
problems that can arise when analyzing large, asymmetric, and potentially
weakly-related collections. Topic correlations in weakly-related collections
typically lie in the tail of the topic distribution, where they would be
overlooked by models unable to fit large numbers of topics. To efficiently
model this long tail for large-scale analysis, our models implement a parallel
sampling algorithm based on the Metropolis-Hastings and alias methods (Yuan et
al., 2015). The models are first evaluated on synthetic data, generated to
simulate various collection-level asymmetries. We then present a case study of
modeling over 300k documents in collections of sciences and humanities research
from JSTOR.
| 2,015 | Computation and Language |
Auto-Sizing Neural Networks: With Applications to n-gram Language Models | Neural networks have been shown to improve performance across a range of
natural-language tasks. However, designing and training them can be
complicated. Frequently, researchers resort to repeated experimentation to pick
optimal settings. In this paper, we address the issue of choosing the correct
number of units in hidden layers. We introduce a method for automatically
adjusting network size by pruning out hidden units through $\ell_{\infty,1}$
and $\ell_{2,1}$ regularization. We apply this method to language modeling and
demonstrate its ability to correctly choose the number of hidden units while
maintaining perplexity. We also include these models in a machine translation
decoder and show that these smaller neural models maintain the significant
improvements of their unpruned versions.
| 2,015 | Computation and Language |
Posterior calibration and exploratory analysis for natural language
processing models | Many models in natural language processing define probabilistic distributions
over linguistic structures. We argue that (1) the quality of a model' s
posterior distribution can and should be directly evaluated, as to whether
probabilities correspond to empirical frequencies, and (2) NLP uncertainty can
be projected not only to pipeline components, but also to exploratory data
analysis, telling a user when to trust and not trust the NLP analysis. We
present a method to analyze calibration, and apply it to compare the
miscalibration of several commonly used models. We also contribute a
coreference sampling algorithm that can create confidence intervals for a
political event extraction task.
| 2,015 | Computation and Language |
Simple Text Mining for Sentiment Analysis of Political Figure Using
Naive Bayes Classifier Method | Text mining can be applied to many fields. One of the application is using
text mining in digital newspaper to do politic sentiment analysis. In this
paper sentiment analysis is applied to get information from digital news
articles about its positive or negative sentiment regarding particular
politician. This paper suggests a simple model to analyze digital newspaper
sentiment polarity using naive Bayes classifier method. The model uses a set of
initial data to begin with which will be updated when new information appears.
The model showed promising result when tested and can be implemented to some
other sentiment analysis problems.
| 2,015 | Computation and Language |
A large annotated corpus for learning natural language inference | Understanding entailment and contradiction is fundamental to understanding
natural language, and inference about entailment and contradiction is a
valuable testing ground for the development of semantic representations.
However, machine learning research in this area has been dramatically limited
by the lack of large-scale resources. To address this, we introduce the
Stanford Natural Language Inference corpus, a new, freely available collection
of labeled sentence pairs, written by humans doing a novel grounded task based
on image captioning. At 570K pairs, it is two orders of magnitude larger than
all other resources of its type. This increase in scale allows lexicalized
classifiers to outperform some sophisticated existing entailment models, and it
allows a neural network-based model to perform competitively on natural
language inference benchmarks for the first time.
| 2,015 | Computation and Language |
Echoes of Persuasion: The Effect of Euphony in Persuasive Communication | While the effect of various lexical, syntactic, semantic and stylistic
features have been addressed in persuasive language from a computational point
of view, the persuasive effect of phonetics has received little attention. By
modeling a notion of euphony and analyzing four datasets comprising persuasive
and non-persuasive sentences in different domains (political speeches, movie
quotes, slogans and tweets), we explore the impact of sounds on different forms
of persuasiveness. We conduct a series of analyses and prediction experiments
within and across datasets. Our results highlight the positive role of phonetic
devices on persuasion.
| 2,015 | Computation and Language |
A Framework for Comparing Groups of Documents | We present a general framework for comparing multiple groups of documents. A
bipartite graph model is proposed where document groups are represented as one
node set and the comparison criteria are represented as the other node set.
Using this model, we present basic algorithms to extract insights into
similarities and differences among the document groups. Finally, we demonstrate
the versatility of our framework through an analysis of NSF funding programs
for basic research.
| 2,015 | Computation and Language |
Better Summarization Evaluation with Word Embeddings for ROUGE | ROUGE is a widely adopted, automatic evaluation measure for text
summarization. While it has been shown to correlate well with human judgements,
it is biased towards surface lexical similarities. This makes it unsuitable for
the evaluation of abstractive summarization, or summaries with substantial
paraphrasing. We study the effectiveness of word embeddings to overcome this
disadvantage of ROUGE. Specifically, instead of measuring lexical overlaps,
word embeddings are used to compute the semantic similarity of the words used
in summaries instead. Our experimental results show that our proposal is able
to achieve better correlations with human judgements when measured with the
Spearman and Kendall rank coefficients.
| 2,015 | Computation and Language |
Visualizing NLP annotations for Crowdsourcing | Visualizing NLP annotation is useful for the collection of training data for
the statistical NLP approaches. Existing toolkits either provide limited visual
aid, or introduce comprehensive operators to realize sophisticated linguistic
rules. Workers must be well trained to use them. Their audience thus can hardly
be scaled to large amounts of non-expert crowdsourced workers. In this paper,
we present CROWDANNO, a visualization toolkit to allow crowd-sourced workers to
annotate two general categories of NLP problems: clustering and parsing.
Workers can finish the tasks with simplified operators in an interactive
interface, and fix errors conveniently. User studies show our toolkit is very
friendly to NLP non-experts, and allow them to produce high quality labels for
several sophisticated problems. We release our source code and toolkit to spur
future research.
| 2,015 | Computation and Language |
A fully data-driven method to identify (correlated) changes in
diachronic corpora | In this paper, a method for measuring synchronic corpus (dis-)similarity put
forward by Kilgarriff (2001) is adapted and extended to identify trends and
correlated changes in diachronic text data, using the Corpus of Historical
American English (Davies 2010a) and the Google Ngram Corpora (Michel et al.
2010a). This paper shows that this fully data-driven method, which extracts
word types that have undergone the most pronounced change in frequency in a
given period of time, is computationally very cheap and that it allows
interpretations of diachronic trends that are both intuitively plausible and
motivated from the perspective of information theory. Furthermore, it
demonstrates that the method is able to identify correlated linguistic changes
and diachronic shifts that can be linked to historical events. Finally, it can
help to improve diachronic POS tagging and complement existing NLP approaches.
This indicates that the approach can facilitate an improved understanding of
diachronic processes in language change.
| 2,015 | Computation and Language |
Crossings as a side effect of dependency lengths | The syntactic structure of sentences exhibits a striking regularity:
dependencies tend to not cross when drawn above the sentence. We investigate
two competing explanations. The traditional hypothesis is that this trend
arises from an independent principle of syntax that reduces crossings
practically to zero. An alternative to this view is the hypothesis that
crossings are a side effect of dependency lengths, i.e. sentences with shorter
dependency lengths should tend to have fewer crossings. We are able to reject
the traditional view in the majority of languages considered. The alternative
hypothesis can lead to a more parsimonious theory of language.
| 2,016 | Computation and Language |
Alignment-based compositional semantics for instruction following | This paper describes an alignment-based model for interpreting natural
language instructions in context. We approach instruction following as a search
over plans, scoring sequences of actions conditioned on structured observations
of text and the environment. By explicitly modeling both the low-level
compositional structure of individual actions and the high-level structure of
full plans, we are able to learn both grounded representations of sentence
meaning and pragmatic constraints on interpretation. To demonstrate the model's
flexibility, we apply it to a diverse set of benchmark tasks. On every task, we
outperform strong task-specific baselines, and achieve several new
state-of-the-art results.
| 2,017 | Computation and Language |
Character-Aware Neural Language Models | We describe a simple neural language model that relies only on
character-level inputs. Predictions are still made at the word-level. Our model
employs a convolutional neural network (CNN) and a highway network over
characters, whose output is given to a long short-term memory (LSTM) recurrent
neural network language model (RNN-LM). On the English Penn Treebank the model
is on par with the existing state-of-the-art despite having 60% fewer
parameters. On languages with rich morphology (Arabic, Czech, French, German,
Spanish, Russian), the model outperforms word-level/morpheme-level LSTM
baselines, again with fewer parameters. The results suggest that on many
languages, character inputs are sufficient for language modeling. Analysis of
word representations obtained from the character composition part of the model
reveals that the model is able to encode, from characters only, both semantic
and orthographic information.
| 2,015 | Computation and Language |
Component-Enhanced Chinese Character Embeddings | Distributed word representations are very useful for capturing semantic
information and have been successfully applied in a variety of NLP tasks,
especially on English. In this work, we innovatively develop two
component-enhanced Chinese character embedding models and their bigram
extensions. Distinguished from English word embeddings, our models explore the
compositions of Chinese characters, which often serve as semantic indictors
inherently. The evaluations on both word similarity and text classification
demonstrate the effectiveness of our models.
| 2,015 | Computation and Language |
Computational Sociolinguistics: A Survey | Language is a social phenomenon and variation is inherent to its social
nature. Recently, there has been a surge of interest within the computational
linguistics (CL) community in the social dimension of language. In this article
we present a survey of the emerging field of "Computational Sociolinguistics"
that reflects this increased interest. We aim to provide a comprehensive
overview of CL research on sociolinguistic themes, featuring topics such as the
relation between language and social identity, language use in social
interaction and multilingual communication. Moreover, we demonstrate the
potential for synergy between the research communities involved, by showing how
the large-scale data-driven methods that are widely used in CL can complement
existing sociolinguistic studies, and how sociolinguistics can inform and
challenge the methods and assumptions employed in CL studies. We hope to convey
the possible benefits of a closer collaboration between the two communities and
conclude with a discussion of open challenges.
| 2,016 | Computation and Language |
An Event Network for Exploring Open Information | In this paper, an event network is presented for exploring open information,
where linguistic units about an event are organized for analysing. The process
is divided into three steps: document event detection, event network
construction and event network analysis. First, by implementing event detection
or tracking, documents are retrospectively (or on-line) organized into document
events. Secondly, for each of the document event, linguistic units are
extracted and combined into event networks. Thirdly, various analytic methods
are proposed for event network analysis. In our application methodologies are
presented for exploring open information.
| 2,015 | Computation and Language |
Word Representations, Tree Models and Syntactic Functions | Word representations induced from models with discrete latent variables
(e.g.\ HMMs) have been shown to be beneficial in many NLP applications. In this
work, we exploit labeled syntactic dependency trees and formalize the induction
problem as unsupervised learning of tree-structured hidden Markov models.
Syntactic functions are used as additional observed variables in the model,
influencing both transition and emission components. Such syntactic information
can potentially lead to capturing more fine-grain and functional distinctions
between words, which, in turn, may be desirable in many NLP applications. We
evaluate the word representations on two tasks -- named entity recognition and
semantic frame identification. We observe improvements from exploiting
syntactic function information in both cases, and the results rivaling those of
state-of-the-art representation learning methods. Additionally, we revisit the
relationship between sequential and unlabeled-tree models and find that the
advantage of the latter is not self-evident.
| 2,016 | Computation and Language |
Neural Machine Translation of Rare Words with Subword Units | Neural machine translation (NMT) models typically operate with a fixed
vocabulary, but translation is an open-vocabulary problem. Previous work
addresses the translation of out-of-vocabulary words by backing off to a
dictionary. In this paper, we introduce a simpler and more effective approach,
making the NMT model capable of open-vocabulary translation by encoding rare
and unknown words as sequences of subword units. This is based on the intuition
that various word classes are translatable via smaller units than words, for
instance names (via character copying or transliteration), compounds (via
compositional translation), and cognates and loanwords (via phonological and
morphological transformations). We discuss the suitability of different word
segmentation techniques, including simple character n-gram models and a
segmentation based on the byte pair encoding compression algorithm, and
empirically show that subword models improve over a back-off dictionary
baseline for the WMT 15 translation tasks English-German and English-Russian by
1.1 and 1.3 BLEU, respectively.
| 2,016 | Computation and Language |
A Neural Attention Model for Abstractive Sentence Summarization | Summarization based on text extraction is inherently limited, but
generation-style abstractive methods have proven challenging to build. In this
work, we propose a fully data-driven approach to abstractive sentence
summarization. Our method utilizes a local attention-based model that generates
each word of the summary conditioned on the input sentence. While the model is
structurally simple, it can easily be trained end-to-end and scales to a large
amount of training data. The model shows significant performance gains on the
DUC-2004 shared task compared with several strong baselines.
| 2,015 | Computation and Language |
Analysis of Communication Pattern with Scammers in Enron Corpus | This paper is an exploratory analysis into fraud detection taking Enron email
corpus as the case study. The paper posits conclusions like strict servitude
and unquestionable faith among employees as breeding grounds for sham among
higher executives. We also try to infer on the nature of communication between
fraudulent employees and between non- fraudulent-fraudulent employees
| 2,015 | Computation and Language |
What to talk about and how? Selective Generation using LSTMs with
Coarse-to-Fine Alignment | We propose an end-to-end, domain-independent neural encoder-aligner-decoder
model for selective generation, i.e., the joint task of content selection and
surface realization. Our model first encodes a full set of over-determined
database event records via an LSTM-based recurrent neural network, then
utilizes a novel coarse-to-fine aligner to identify the small subset of salient
records to talk about, and finally employs a decoder to generate free-form
descriptions of the aligned, selected records. Our model achieves the best
selection and generation results reported to-date (with 59% relative
improvement in generation) on the benchmark WeatherGov dataset, despite using
no specialized features or linguistic resources. Using an improved k-nearest
neighbor beam filter helps further. We also perform a series of ablations and
visualizations to elucidate the contributions of our key model components.
Lastly, we evaluate the generalizability of our model on the RoboCup dataset,
and get results that are competitive with or better than the state-of-the-art,
despite being severely data-starved.
| 2,016 | Computation and Language |
On TimeML-Compliant Temporal Expression Extraction in Turkish | It is commonly acknowledged that temporal expression extractors are important
components of larger natural language processing systems like information
retrieval and question answering systems. Extraction and normalization of
temporal expressions in Turkish has not been given attention so far except the
extraction of some date and time expressions within the course of named entity
recognition. As TimeML is the current standard of temporal expression and event
annotation in natural language texts, in this paper, we present an analysis of
temporal expressions in Turkish based on the related TimeML classification
(i.e., date, time, duration, and set expressions). We have created a lexicon
for Turkish temporal expressions and devised considerably wide-coverage
patterns using the lexical classes as the building blocks. We believe that the
proposed patterns, together with convenient normalization rules, can be readily
used by prospective temporal expression extraction tools for Turkish.
| 2,015 | Computation and Language |
Encoding Prior Knowledge with Eigenword Embeddings | Canonical correlation analysis (CCA) is a method for reducing the dimension
of data represented using two views. It has been previously used to derive word
embeddings, where one view indicates a word, and the other view indicates its
context. We describe a way to incorporate prior knowledge into CCA, give a
theoretical justification for it, and test it by deriving word embeddings and
evaluating them on a myriad of datasets.
| 2,016 | Computation and Language |
The influence of Chunking on Dependency Crossing and Distance | This paper hypothesizes that chunking plays important role in reducing
dependency distance and dependency crossings. Computer simulations, when
compared with natural languages,show that chunking reduces mean dependency
distance (MDD) of a linear sequence of nodes (constrained by continuity or
projectivity) to that of natural languages. More interestingly, chunking alone
brings about less dependency crossings as well, though having failed to reduce
them, to such rarity as found in human languages. These results suggest that
chunking may play a vital role in the minimization of dependency distance, and
a somewhat contributing role in the rarity of dependency crossing. In addition,
the results point to a possibility that the rarity of dependency crossings is
not a mere side-effect of minimization of dependency distance, but a linguistic
phenomenon with its own motivations.
| 2,016 | Computation and Language |
Better Document-level Sentiment Analysis from RST Discourse Parsing | Discourse structure is the hidden link between surface features and
document-level properties, such as sentiment polarity. We show that the
discourse analyses produced by Rhetorical Structure Theory (RST) parsers can
improve document-level sentiment analysis, via composition of local information
up the discourse tree. First, we show that reweighting discourse units
according to their position in a dependency representation of the rhetorical
structure can yield substantial improvements on lexicon-based sentiment
analysis. Next, we present a recursive neural network over the RST structure,
which offers significant improvements over classification-based methods.
| 2,015 | Computation and Language |
Take and Took, Gaggle and Goose, Book and Read: Evaluating the Utility
of Vector Differences for Lexical Relation Learning | Recent work on word embeddings has shown that simple vector subtraction over
pre-trained embeddings is surprisingly effective at capturing different lexical
relations, despite lacking explicit supervision. Prior work has evaluated this
intriguing result using a word analogy prediction formulation and hand-selected
relations, but the generality of the finding over a broader range of lexical
relation types and different learning settings has not been evaluated. In this
paper, we carry out such an evaluation in two learning settings: (1) spectral
clustering to induce word relations, and (2) supervised learning to classify
vector differences into relation types. We find that word embeddings capture a
surprising amount of information, and that, under suitable supervised training,
vector subtraction generalises well to a broad range of relations, including
over unseen lexical items.
| 2,016 | Computation and Language |
A commentary on "The now-or-never bottleneck: a fundamental constraint
on language", by Christiansen and Chater (2016) | In a recent article, Christiansen and Chater (2016) present a fundamental
constraint on language, i.e. a now-or-never bottleneck that arises from our
fleeting memory, and explore its implications, e.g., chunk-and-pass processing,
outlining a framework that promises to unify different areas of research. Here
we explore additional support for this constraint and suggest further
connections from quantitative linguistics and information theory.
| 2,017 | Computation and Language |
Integrate Document Ranking Information into Confidence Measure
Calculation for Spoken Term Detection | This paper proposes an algorithm to improve the calculation of confidence
measure for spoken term detection (STD). Given an input query term, the
algorithm first calculates a measurement named document ranking weight for each
document in the speech database to reflect its relevance with the query term by
summing all the confidence measures of the hypothesized term occurrences in
this document. The confidence measure of each term occurrence is then
re-estimated through linear interpolation with the calculated document ranking
weight to improve its reliability by integrating document-level information.
Experiments are conducted on three standard STD tasks for Tamil, Vietnamese and
English respectively. The experimental results all demonstrate that the
proposed algorithm achieves consistent improvements over the state-of-the-art
method for confidence measure calculation. Furthermore, this algorithm is still
effective even if a high accuracy speech recognizer is not available, which
makes it applicable for the languages with limited speech resources.
| 2,015 | Computation and Language |
Exploiting Out-of-Domain Data Sources for Dialectal Arabic Statistical
Machine Translation | Statistical machine translation for dialectal Arabic is characterized by a
lack of data since data acquisition involves the transcription and translation
of spoken language. In this study we develop techniques for extracting parallel
data for one particular dialect of Arabic (Iraqi Arabic) from out-of-domain
corpora in different dialects of Arabic or in Modern Standard Arabic. We
compare two different data selection strategies (cross-entropy based and
submodular selection) and demonstrate that a very small but highly targeted
amount of found data can improve the performance of a baseline machine
translation system. We furthermore report on preliminary experiments on using
automatically translated speech data as additional training data.
| 2,015 | Computation and Language |
Unsupervised Discovery of Linguistic Structure Including Two-level
Acoustic Patterns Using Three Cascaded Stages of Iterative Optimization | Techniques for unsupervised discovery of acoustic patterns are getting
increasingly attractive, because huge quantities of speech data are becoming
available but manual annotations remain hard to acquire. In this paper, we
propose an approach for unsupervised discovery of linguistic structure for the
target spoken language given raw speech data. This linguistic structure
includes two-level (subword-like and word-like) acoustic patterns, the lexicon
of word-like patterns in terms of subword-like patterns and the N-gram language
model based on word-like patterns. All patterns, models, and parameters can be
automatically learned from the unlabelled speech corpus. This is achieved by an
initialization step followed by three cascaded stages for acoustic, linguistic,
and lexical iterative optimization. The lexicon of word-like patterns defines
allowed consecutive sequence of HMMs for subword-like patterns. In each
iteration, model training and decoding produces updated labels from which the
lexicon and HMMs can be further updated. In this way, model parameters and
decoded labels are respectively optimized in each iteration, and the knowledge
about the linguistic structure is learned gradually layer after layer. The
proposed approach was tested in preliminary experiments on a corpus of Mandarin
broadcast news, including a task of spoken term detection with performance
compared to a parallel test using models trained in a supervised way. Results
show that the proposed system not only yields reasonable performance on its
own, but is also complimentary to existing large vocabulary ASR systems.
| 2,015 | Computation and Language |
Unsupervised Spoken Term Detection with Spoken Queries by Multi-level
Acoustic Patterns with Varying Model Granularity | This paper presents a new approach for unsupervised Spoken Term Detection
with spoken queries using multiple sets of acoustic patterns automatically
discovered from the target corpus. The different pattern HMM
configurations(number of states per model, number of distinct models, number of
Gaussians per state)form a three-dimensional model granularity space. Different
sets of acoustic patterns automatically discovered on different points properly
distributed over this three-dimensional space are complementary to one another,
thus can jointly capture the characteristics of the spoken terms. By
representing the spoken content and spoken query as sequences of acoustic
patterns, a series of approaches for matching the pattern index sequences while
considering the signal variations are developed. In this way, not only the
on-line computation load can be reduced, but the signal distributions caused by
different speakers and acoustic conditions can be reasonably taken care of. The
results indicate that this approach significantly outperformed the unsupervised
feature-based DTW baseline by 16.16\% in mean average precision on the TIMIT
corpus.
| 2,015 | Computation and Language |
Enhancing Automatically Discovered Multi-level Acoustic Patterns
Considering Context Consistency With Applications in Spoken Term Detection | This paper presents a novel approach for enhancing the multiple sets of
acoustic patterns automatically discovered from a given corpus. In a previous
work it was proposed that different HMM configurations (number of states per
model, number of distinct models) for the acoustic patterns form a
two-dimensional space. Multiple sets of acoustic patterns automatically
discovered with the HMM configurations properly located on different points
over this two-dimensional space were shown to be complementary to one another,
jointly capturing the characteristics of the given corpus. By representing the
given corpus as sequences of acoustic patterns on different HMM sets, the
pattern indices in these sequences can be relabeled considering the context
consistency across the different sequences. Good improvements were observed in
preliminary experiments of pattern spoken term detection (STD) performed on
both TIMIT and Mandarin Broadcast News with such enhanced patterns.
| 2,015 | Computation and Language |
Probabilistic Bag-Of-Hyperlinks Model for Entity Linking | Many fundamental problems in natural language processing rely on determining
what entities appear in a given text. Commonly referenced as entity linking,
this step is a fundamental component of many NLP tasks such as text
understanding, automatic summarization, semantic search or machine translation.
Name ambiguity, word polysemy, context dependencies and a heavy-tailed
distribution of entities contribute to the complexity of this problem.
We here propose a probabilistic approach that makes use of an effective
graphical model to perform collective entity disambiguation. Input mentions
(i.e.,~linkable token spans) are disambiguated jointly across an entire
document by combining a document-level prior of entity co-occurrences with
local information captured from mentions and their surrounding context. The
model is based on simple sufficient statistics extracted from data, thus
relying on few parameters to be learned.
Our method does not require extensive feature engineering, nor an expensive
training procedure. We use loopy belief propagation to perform approximate
inference. The low complexity of our model makes this step sufficiently fast
for real-time usage. We demonstrate the accuracy of our approach on a wide
range of benchmark datasets, showing that it matches, and in many cases
outperforms, existing state-of-the-art methods.
| 2,016 | Computation and Language |
Unsupervised Domain Discovery using Latent Dirichlet Allocation for
Acoustic Modelling in Speech Recognition | Speech recognition systems are often highly domain dependent, a fact widely
reported in the literature. However the concept of domain is complex and not
bound to clear criteria. Hence it is often not evident if data should be
considered to be out-of-domain. While both acoustic and language models can be
domain specific, work in this paper concentrates on acoustic modelling. We
present a novel method to perform unsupervised discovery of domains using
Latent Dirichlet Allocation (LDA) modelling. Here a set of hidden domains is
assumed to exist in the data, whereby each audio segment can be considered to
be a weighted mixture of domain properties. The classification of audio
segments into domains allows the creation of domain specific acoustic models
for automatic speech recognition. Experiments are conducted on a dataset of
diverse speech data covering speech from radio and TV broadcasts, telephone
conversations, meetings, lectures and read speech, with a joint training set of
60 hours and a test set of 6 hours. Maximum A Posteriori (MAP) adaptation to
LDA based domains was shown to yield relative Word Error Rate (WER)
improvements of up to 16% relative, compared to pooled training, and up to 10%,
compared with models adapted with human-labelled prior domain knowledge.
| 2,015 | Computation and Language |
Towards Understanding Egyptian Arabic Dialogues | Labelling of user's utterances to understanding his attends which called
Dialogue Act (DA) classification, it is considered the key player for dialogue
language understanding layer in automatic dialogue systems. In this paper, we
proposed a novel approach to user's utterances labeling for Egyptian
spontaneous dialogues and Instant Messages using Machine Learning (ML) approach
without relying on any special lexicons, cues, or rules. Due to the lack of
Egyptian dialect dialogue corpus, the system evaluated by multi-genre corpus
includes 4725 utterances for three domains, which are collected and annotated
manually from Egyptian call-centers. The system achieves F1 scores of 70. 36%
overall domains.
| 2,015 | Computation and Language |
Liberating language research from dogmas of the 20th century | A commentary on the article "Large-scale evidence of dependency length
minimization in 37 languages" by Futrell, Mahowald & Gibson (PNAS 2015 112 (33)
10336-10341).
| 2,016 | Computation and Language |
Verbs Taking Clausal and Non-Finite Arguments as Signals of Modality -
Revisiting the Issue of Meaning Grounded in Syntax | We revisit Levin's theory about the correspondence of verb meaning and syntax
and infer semantic classes from a large syntactic classification of more than
600 German verbs taking clausal and non-finite arguments. Grasping the meaning
components of Levin-classes is known to be hard. We address this challenge by
setting up a multi-perspective semantic characterization of the inferred
classes. To this end, we link the inferred classes and their English
translation to independently constructed semantic classes in three different
lexicons - the German wordnet GermaNet, VerbNet and FrameNet - and perform a
detailed analysis and evaluation of the resulting German-English classification
(available at www.ukp.tu-darmstadt.de/modality-verbclasses/).
| 2,016 | Computation and Language |
A Parallel Corpus of Translationese | We describe a set of bilingual English--French and English--German parallel
corpora in which the direction of translation is accurately and reliably
annotated. The corpora are diverse, consisting of parliamentary proceedings,
literary works, transcriptions of TED talks and political commentary. They will
be instrumental for research of translationese and its applications to (human
and machine) translation; specifically, they can be used for the task of
translationese identification, a research direction that enjoys a growing
interest in recent years. To validate the quality and reliability of the
corpora, we replicated previous results of supervised and unsupervised
identification of translationese, and further extended the experiments to
additional datasets and languages.
| 2,016 | Computation and Language |
Improving distant supervision using inference learning | Distant supervision is a widely applied approach to automatic training of
relation extraction systems and has the advantage that it can generate large
amounts of labelled data with minimal effort. However, this data may contain
errors and consequently systems trained using distant supervision tend not to
perform as well as those based on manually labelled data. This work proposes a
novel method for detecting potential false negative training examples using a
knowledge inference method. Results show that our approach improves the
performance of relation extraction systems trained using distantly supervised
data.
| 2,015 | Computation and Language |
The USFD Spoken Language Translation System for IWSLT 2014 | The University of Sheffield (USFD) participated in the International Workshop
for Spoken Language Translation (IWSLT) in 2014. In this paper, we will
introduce the USFD SLT system for IWSLT. Automatic speech recognition (ASR) is
achieved by two multi-pass deep neural network systems with adaptation and
rescoring techniques. Machine translation (MT) is achieved by a phrase-based
system. The USFD primary system incorporates state-of-the-art ASR and MT
techniques and gives a BLEU score of 23.45 and 14.75 on the English-to-French
and English-to-German speech-to-text translation task with the IWSLT 2014 data.
The USFD contrastive systems explore the integration of ASR and MT by using a
quality estimation system to rescore the ASR outputs, optimising towards better
translation. This gives a further 0.54 and 0.26 BLEU improvement respectively
on the IWSLT 2012 and 2014 evaluation data.
| 2,014 | Computation and Language |
Twitter Sentiment Analysis | This project addresses the problem of sentiment analysis in twitter; that is
classifying tweets according to the sentiment expressed in them: positive,
negative or neutral. Twitter is an online micro-blogging and social-networking
platform which allows users to write short status updates of maximum length 140
characters. It is a rapidly expanding service with over 200 million registered
users - out of which 100 million are active users and half of them log on
twitter on a daily basis - generating nearly 250 million tweets per day. Due to
this large amount of usage we hope to achieve a reflection of public sentiment
by analysing the sentiments expressed in the tweets. Analysing the public
sentiment is important for many applications such as firms trying to find out
the response of their products in the market, predicting political elections
and predicting socioeconomic phenomena like stock exchange. The aim of this
project is to develop a functional classifier for accurate and automatic
sentiment classification of an unknown tweet stream.
| 2,015 | Computation and Language |
Kannada named entity recognition and classification (nerc) based on
multinomial na\"ive bayes (mnb) classifier | Named Entity Recognition and Classification (NERC) is a process of
identification of proper nouns in the text and classification of those nouns
into certain predefined categories like person name, location, organization,
date, and time etc. NERC in Kannada is an essential and challenging task. The
aim of this work is to develop a novel model for NERC, based on Multinomial
Na\"ive Bayes (MNB) Classifier. The Methodology adopted in this paper is based
on feature extraction of training corpus, by using term frequency, inverse
document frequency and fitting them to a tf-idf-vectorizer. The paper discusses
the various issues in developing the proposed model. The details of
implementation and performance evaluation are discussed. The experiments are
conducted on a training corpus of size 95,170 tokens and test corpus of 5,000
tokens. It is observed that the model works with Precision, Recall and
F1-measure of 83%, 79% and 81% respectively.
| 2,015 | Computation and Language |
Dependency length minimization: Puzzles and Promises | In the recent issue of PNAS, Futrell et al. claims that their study of 37
languages gives the first large scale cross-language evidence for Dependency
Length Minimization, which is an overstatement that ignores similar previous
researches. In addition,this study seems to pay no attention to factors like
the uniformity of genres,which weakens the validity of the argument that DLM is
universal. Another problem is that this study sets the baseline random language
as projective, which fails to truly uncover the difference between natural
language and random language, since projectivity is an important feature of
many natural languages. Finally, the paper contends an "apparent relationship
between head finality and dependency length" despite the lack of an explicit
statistical comparison, which renders this conclusion rather hasty and
improper.
| 2,016 | Computation and Language |
Splitting Compounds by Semantic Analogy | Compounding is a highly productive word-formation process in some languages
that is often problematic for natural language processing applications. In this
paper, we investigate whether distributional semantics in the form of word
embeddings can enable a deeper, i.e., more knowledge-rich, processing of
compounds than the standard string-based methods. We present an unsupervised
approach that exploits regularities in the semantic vector space (based on
analogies such as "bookshop is to shop as bookshelf is to shelf") to produce
compound analyses of high quality. A subsequent compound splitting algorithm
based on these analyses is highly effective, particularly for ambiguous
compounds. German to English machine translation experiments show that this
semantic analogy-based compound splitter leads to better translations than a
commonly used frequency-based method.
| 2,015 | Computation and Language |
amLite: Amharic Transliteration Using Key Map Dictionary | amLite is a framework developed to map ASCII transliterated Amharic texts
back to the original Amharic letter texts. The aim of such a framework is to
make existing Amharic linguistic data consistent and interoperable among
researchers. For achieving the objective, a key map dictionary is constructed
using the possible ASCII combinations actively in use for transliterating
Amharic letters; and a mapping of the combinations to the corresponding Amharic
letters is done. The mapping is then used to replace the Amharic linguistic
text back to form the original Amharic letters text. The framework indicated
97.7, 99.7 and 98.4 percentage accuracy on converting the three sample random
test data. It is; however, possible to improve the accuracy of the framework by
adding an exception to the implementation of the algorithm, or by preprocessing
the input text prior to conversion. This paper outlined the rationales behind
the need for developing the framework and the processes undertaken in the
development.
| 2,015 | Computation and Language |
Extraction of evidence tables from abstracts of randomized clinical
trials using a maximum entropy classifier and global constraints | Systematic use of the published results of randomized clinical trials is
increasingly important in evidence-based medicine. In order to collate and
analyze the results from potentially numerous trials, evidence tables are used
to represent trials concerning a set of interventions of interest. An evidence
table has columns for the patient group, for each of the interventions being
compared, for the criterion for the comparison (e.g. proportion who survived
after 5 years from treatment), and for each of the results. Currently, it is a
labour-intensive activity to read each published paper and extract the
information for each field in an evidence table. There have been some NLP
studies investigating how some of the features from papers can be extracted, or
at least the relevant sentences identified. However, there is a lack of an NLP
system for the systematic extraction of each item of information required for
an evidence table. We address this need by a combination of a maximum entropy
classifier, and integer linear programming. We use the later to handle
constraints on what is an acceptable classification of the features to be
extracted. With experimental results, we demonstrate substantial advantages in
using global constraints (such as the features describing the patient group,
and the interventions, must occur before the features describing the results of
the comparison).
| 2,015 | Computation and Language |
Network analysis of named entity co-occurrences in written texts | The use of methods borrowed from statistics and physics to analyze written
texts has allowed the discovery of unprecedent patterns of human behavior and
cognition by establishing links between models features and language structure.
While current models have been useful to unveil patterns via analysis of
syntactical and semantical networks, only a few works have probed the relevance
of investigating the structure arising from the relationship between relevant
entities such as characters, locations and organizations. In this study, we
represent entities appearing in the same context as a co-occurrence network,
where links are established according to a null model based on random, shuffled
texts. Computational simulations performed in novels revealed that the proposed
model displays interesting topological features, such as the small world
feature, characterized by high values of clustering coefficient. The
effectiveness of our model was verified in a practical pattern recognition task
in real networks. When compared with traditional word adjacency networks, our
model displayed optimized results in identifying unknown references in texts.
Because the proposed representation plays a complementary role in
characterizing unstructured documents via topological analysis of named
entities, we believe that it could be useful to improve the characterization of
written texts (and related systems), specially if combined with traditional
approaches based on statistical and deeper paradigms.
| 2,016 | Computation and Language |
TransG : A Generative Mixture Model for Knowledge Graph Embedding | Recently, knowledge graph embedding, which projects symbolic entities and
relations into continuous vector space, has become a new, hot topic in
artificial intelligence. This paper addresses a new issue of multiple relation
semantics that a relation may have multiple meanings revealed by the entity
pairs associated with the corresponding triples, and proposes a novel Gaussian
mixture model for embedding, TransG. The new model can discover latent
semantics for a relation and leverage a mixture of relation component vectors
for embedding a fact triple. To the best of our knowledge, this is the first
generative model for knowledge graph embedding, which is able to deal with
multiple relation semantics. Extensive experiments show that the proposed model
achieves substantial improvements against the state-of-the-art baselines.
| 2,017 | Computation and Language |
TransA: An Adaptive Approach for Knowledge Graph Embedding | Knowledge representation is a major topic in AI, and many studies attempt to
represent entities and relations of knowledge base in a continuous vector
space. Among these attempts, translation-based methods build entity and
relation vectors by minimizing the translation loss from a head entity to a
tail one. In spite of the success of these methods, translation-based methods
also suffer from the oversimplified loss metric, and are not competitive enough
to model various and complex entities/relations in knowledge bases. To address
this issue, we propose \textbf{TransA}, an adaptive metric approach for
embedding, utilizing the metric learning ideas to provide a more flexible
embedding method. Experiments are conducted on the benchmark datasets and our
proposed method makes significant and consistent improvements over the
state-of-the-art baselines.
| 2,015 | Computation and Language |
A Light Sliding-Window Part-of-Speech Tagger for the Apertium
Free/Open-Source Machine Translation Platform | This paper describes a free/open-source implementation of the light
sliding-window (LSW) part-of-speech tagger for the Apertium free/open-source
machine translation platform. Firstly, the mechanism and training process of
the tagger are reviewed, and a new method for incorporating linguistic rules is
proposed. Secondly, experiments are conducted to compare the performances of
the tagger under different window settings, with or without Apertium-style
"forbid" rules, with or without Constraint Grammar, and also with respect to
the traditional HMM tagger in Apertium.
| 2,015 | Computation and Language |
Word, graph and manifold embedding from Markov processes | Continuous vector representations of words and objects appear to carry
surprisingly rich semantic content. In this paper, we advance both the
conceptual and theoretical understanding of word embeddings in three ways.
First, we ground embeddings in semantic spaces studied in
cognitive-psychometric literature and introduce new evaluation tasks. Second,
in contrast to prior work, we take metric recovery as the key object of study,
unify existing algorithms as consistent metric recovery methods based on
co-occurrence counts from simple Markov random walks, and propose a new
recovery algorithm. Third, we generalize metric recovery to graphs and
manifolds, relating co-occurence counts on random walks in graphs and random
processes on manifolds to the underlying metric to be recovered, thereby
reconciling manifold estimation and embedding algorithms. We compare embedding
algorithms across a range of tasks, from nonlinear dimensionality reduction to
three semantic language tasks, including analogies, sequence completion, and
classification.
| 2,015 | Computation and Language |
Early text classification: a Naive solution | Text classification is a widely studied problem, and it can be considered
solved for some domains and under certain circumstances. There are scenarios,
however, that have received little or no attention at all, despite its
relevance and applicability. One of such scenarios is early text
classification, where one needs to know the category of a document by using
partial information only. A document is processed as a sequence of terms, and
the goal is to devise a method that can make predictions as fast as possible.
The importance of this variant of the text classification problem is evident in
domains like sexual predator detection, where one wants to identify an offender
as early as possible. This paper analyzes the suitability of the standard naive
Bayes classifier for approaching this problem. Specifically, we assess its
performance when classifying documents after seeing an increasingly number of
terms. A simple modification to the standard naive Bayes implementation allows
us to make predictions with partial information. To the best of our knowledge
naive Bayes has not been used for this purpose before. Throughout an extensive
experimental evaluation we show the effectiveness of the classifier for early
text classification. What is more, we show that this simple solution is very
competitive when compared with state of the art methodologies that are more
elaborated. We foresee our work will pave the way for the development of more
effective early text classification techniques based in the naive Bayes
formulation.
| 2,015 | Computation and Language |
A Review of Features for the Discrimination of Twitter Users:
Application to the Prediction of Offline Influence | Many works related to Twitter aim at characterizing its users in some way:
role on the service (spammers, bots, organizations, etc.), nature of the user
(socio-professional category, age, etc.), topics of interest , and others.
However, for a given user classification problem, it is very difficult to
select a set of appropriate features, because the many features described in
the literature are very heterogeneous, with name overlaps and collisions, and
numerous very close variants. In this article, we review a wide range of such
features. In order to present a clear state-of-the-art description, we unify
their names, definitions and relationships, and we propose a new, neutral,
typology. We then illustrate the interest of our review by applying a selection
of these features to the offline influence detection problem. This task
consists in identifying users which are influential in real-life, based on
their Twitter account and related data. We show that most features deemed
efficient to predict online influence, such as the numbers of retweets and
followers, are not relevant to this problem. However, We propose several
content-based approaches to label Twitter users as Influencers or not. We also
rank them according to a predicted influence level. Our proposals are evaluated
over the CLEF RepLab 2014 dataset, and outmatch state-of-the-art methods.
| 2,016 | Computation and Language |
Reasoning about Entailment with Neural Attention | While most approaches to automatically recognizing entailment relations have
used classifiers employing hand engineered features derived from complex
natural language processing pipelines, in practice their performance has been
only slightly better than bag-of-word pair classifiers using only lexical
similarity. The only attempt so far to build an end-to-end differentiable
neural network for entailment failed to outperform such a simple similarity
classifier. In this paper, we propose a neural model that reads two sentences
to determine entailment using long short-term memory units. We extend this
model with a word-by-word neural attention mechanism that encourages reasoning
over entailments of pairs of words and phrases. Furthermore, we present a
qualitative analysis of attention weights produced by this model, demonstrating
such reasoning capabilities. On a large entailment dataset this model
outperforms the previous best neural model and a classifier with engineered
features by a substantial margin. It is the first generic end-to-end
differentiable system that achieves state-of-the-art accuracy on a textual
entailment dataset.
| 2,016 | Computation and Language |
Automatic Dialect Detection in Arabic Broadcast Speech | We investigate different approaches for dialect identification in Arabic
broadcast speech, using phonetic, lexical features obtained from a speech
recognition system, and acoustic features using the i-vector framework. We
studied both generative and discriminate classifiers, and we combined these
features using a multi-class Support Vector Machine (SVM). We validated our
results on an Arabic/English language identification task, with an accuracy of
100%. We used these features in a binary classifier to discriminate between
Modern Standard Arabic (MSA) and Dialectal Arabic, with an accuracy of 100%. We
further report results using the proposed method to discriminate between the
five most widely used dialects of Arabic: namely Egyptian, Gulf, Levantine,
North African, and MSA, with an accuracy of 52%. We discuss dialect
identification errors in the context of dialect code-switching between
Dialectal Arabic and MSA, and compare the error pattern between manually
labeled data, and the output from our classifier. We also release the train and
test data as standard corpus for dialect identification.
| 2,016 | Computation and Language |
Fully automatic multi-language translation with a catalogue of phrases -
successful employment for the Swiss avalanche bulletin | The Swiss avalanche bulletin is produced twice a day in four languages. Due
to the lack of time available for manual translation, a fully automated
translation system is employed, based on a catalogue of predefined phrases and
predetermined rules of how these phrases can be combined to produce sentences.
Because this catalogue of phrases is limited to a small sublanguage, the system
is able to automatically translate such sentences from German into the target
languages French, Italian and English without subsequent proofreading or
correction. Having been operational for two winter seasons, we assess here the
quality of the produced texts based on two different surveys where participants
rated texts from real avalanche bulletins from both origins, the catalogue of
phrases versus manually written and translated texts. With a mean recognition
rate of 55%, users can hardly distinguish between thetwo types of texts, and
give very similar ratings with respect to their language quality. Overall, the
output from the catalogue system can be considered virtually equivalent to a
text written by avalanche forecasters and then manually translated by
professional translators. Furthermore, forecasters declared that all relevant
situations were captured by the system with sufficient accuracy. Forecaster's
working load did not change with the introduction of the catalogue: the extra
time to find matching sentences is compensated by the fact that they no longer
need to double-check manually translated texts. The reduction of daily
translation costs is expected to offset the initial development costs within a
few years.
| 2,015 | Computation and Language |
Exploration and Exploitation of Victorian Science in Darwin's Reading
Notebooks | Search in an environment with an uncertain distribution of resources involves
a trade-off between exploitation of past discoveries and further exploration.
This extends to information foraging, where a knowledge-seeker shifts between
reading in depth and studying new domains. To study this decision-making
process, we examine the reading choices made by one of the most celebrated
scientists of the modern era: Charles Darwin. From the full-text of books
listed in his chronologically-organized reading journals, we generate topic
models to quantify his local (text-to-text) and global (text-to-past) reading
decisions using Kullback-Liebler Divergence, a cognitively-validated,
information-theoretic measure of relative surprise. Rather than a pattern of
surprise-minimization, corresponding to a pure exploitation strategy, Darwin's
behavior shifts from early exploitation to later exploration, seeking unusually
high levels of cognitive surprise relative to previous eras. These shifts,
detected by an unsupervised Bayesian model, correlate with major intellectual
epochs of his career as identified both by qualitative scholarship and Darwin's
own self-commentary. Our methods allow us to compare his consumption of texts
with their publication order. We find Darwin's consumption more exploratory
than the culture's production, suggesting that underneath gradual societal
changes are the explorations of individual synthesis and discovery. Our
quantitative methods advance the study of cognitive search through a framework
for testing interactions between individual and collective behavior and between
short- and long-term consumption choices. This novel application of topic
modeling to characterize individual reading complements widespread studies of
collective scientific behavior.
| 2,017 | Computation and Language |
Bilingual Distributed Word Representations from Document-Aligned
Comparable Data | We propose a new model for learning bilingual word representations from
non-parallel document-aligned data. Following the recent advances in word
representation learning, our model learns dense real-valued word vectors, that
is, bilingual word embeddings (BWEs). Unlike prior work on inducing BWEs which
heavily relied on parallel sentence-aligned corpora and/or readily available
translation resources such as dictionaries, the article reveals that BWEs may
be learned solely on the basis of document-aligned comparable data without any
additional lexical resources nor syntactic information. We present a comparison
of our approach with previous state-of-the-art models for learning bilingual
word representations from comparable data that rely on the framework of
multilingual probabilistic topic modeling (MuPTM), as well as with
distributional local context-counting models. We demonstrate the utility of the
induced BWEs in two semantic tasks: (1) bilingual lexicon extraction, (2)
suggesting word translations in context for polysemous words. Our simple yet
effective BWE-based models significantly outperform the MuPTM-based and
context-counting representation models from comparable data as well as prior
BWE-based models, and acquire the best reported results on both tasks for all
three tested language pairs.
| 2,016 | Computation and Language |
Description of the Odin Event Extraction Framework and Rule Language | This document describes the Odin framework, which is a domain-independent
platform for developing rule-based event extraction models. Odin aims to be
powerful (the rule language allows the modeling of complex syntactic
structures) and robust (to recover from syntactic parsing errors, syntactic
patterns can be freely mixed with surface, token-based patterns), while
remaining simple (some domain grammars can be up and running in minutes), and
fast (Odin processes over 100 sentences/second in a real-world domain with over
200 rules). Here we include a thorough definition of the Odin rule language,
together with a description of the Odin API in the Scala language, which allows
one to apply these rules to arbitrary texts.
| 2,015 | Computation and Language |
Sentiment Uncertainty and Spam in Twitter Streams and Its Implications
for General Purpose Realtime Sentiment Analysis | State of the art benchmarks for Twitter Sentiment Analysis do not consider
the fact that for more than half of the tweets from the public stream a
distinct sentiment cannot be chosen. This paper provides a new perspective on
Twitter Sentiment Analysis by highlighting the necessity of explicitly
incorporating uncertainty. Moreover, a dataset of high quality to evaluate
solutions for this new problem is introduced and made publicly available.
| 2,015 | Computation and Language |
Sentiment of Emojis | There is a new generation of emoticons, called emojis, that is increasingly
being used in mobile communications and social media. In the past two years,
over ten billion emojis were used on Twitter. Emojis are Unicode graphic
symbols, used as a shorthand to express concepts and ideas. In contrast to the
small number of well-known emoticons that carry clear emotional contents, there
are hundreds of emojis. But what are their emotional contents? We provide the
first emoji sentiment lexicon, called the Emoji Sentiment Ranking, and draw a
sentiment map of the 751 most frequently used emojis. The sentiment of the
emojis is computed from the sentiment of the tweets in which they occur. We
engaged 83 human annotators to label over 1.6 million tweets in 13 European
languages by the sentiment polarity (negative, neutral, or positive). About 4%
of the annotated tweets contain emojis. The sentiment analysis of the emojis
allows us to draw several interesting conclusions. It turns out that most of
the emojis are positive, especially the most popular ones. The sentiment
distribution of the tweets with and without emojis is significantly different.
The inter-annotator agreement on the tweets with emojis is higher. Emojis tend
to occur at the end of the tweets, and their sentiment polarity increases with
the distance. We observe no significant differences in the emoji rankings
between the 13 languages and the Emoji Sentiment Ranking. Consequently, we
propose our Emoji Sentiment Ranking as a European language-independent resource
for automated sentiment analysis. Finally, the paper provides a formalization
of sentiment and a novel visualization in the form of a sentiment bar.
| 2,015 | Computation and Language |
Tuned and GPU-accelerated parallel data mining from comparable corpora | The multilingual nature of the world makes translation a crucial requirement
today. Parallel dictionaries constructed by humans are a widely-available
resource, but they are limited and do not provide enough coverage for good
quality translation purposes, due to out-of-vocabulary words and neologisms.
This motivates the use of statistical translation systems, which are
unfortunately dependent on the quantity and quality of training data. Such has
a very limited availability especially for some languages and very narrow text
domains. Is this research we present our improvements to Yalign mining
methodology by reimplementing the comparison algorithm, introducing a tuning
scripts and by improving performance using GPU computing acceleration. The
experiments are conducted on various text domains and bi-data is extracted from
the Wikipedia dumps.
| 2,015 | Computation and Language |
Neural-based machine translation for medical text domain. Based on
European Medicines Agency leaflet texts | The quality of machine translation is rapidly evolving. Today one can find
several machine translation systems on the web that provide reasonable
translations, although the systems are not perfect. In some specific domains,
the quality may decrease. A recently proposed approach to this domain is neural
machine translation. It aims at building a jointly-tuned single neural network
that maximizes translation performance, a very different approach from
traditional statistical machine translation. Recently proposed neural machine
translation models often belong to the encoder-decoder family in which a source
sentence is encoded into a fixed length vector that is, in turn, decoded to
generate a translation. The present research examines the effects of different
training methods on a Polish-English Machine Translation system used for
medical data. The European Medicines Agency parallel text corpus was used as
the basis for training of neural and statistical network-based translation
systems. The main machine translation evaluation metrics have also been used in
analysis of the systems. A comparison and implementation of a real-time medical
translator is the main focus of our experiments.
| 2,015 | Computation and Language |
Automatically Segmenting Oral History Transcripts | Dividing oral histories into topically coherent segments can make them more
accessible online. People regularly make judgments about where coherent
segments can be extracted from oral histories. But making these judgments can
be taxing, so automated assistance is potentially attractive to speed the task
of extracting segments from open-ended interviews. When different people are
asked to extract coherent segments from the same oral histories, they often do
not agree about precisely where such segments begin and end. This low agreement
makes the evaluation of algorithmic segmenters challenging, but there is reason
to believe that for segmenting oral history transcripts, some approaches are
more promising than others. The BayesSeg algorithm performs slightly better
than TextTiling, while TextTiling does not perform significantly better than a
uniform segmentation. BayesSeg might be used to suggest boundaries to someone
segmenting oral histories, but this segmentation task needs to be better
defined.
| 2,015 | Computation and Language |
Polish - English Speech Statistical Machine Translation Systems for the
IWSLT 2014 | This research explores effects of various training settings between Polish
and English Statistical Machine Translation systems for spoken language.
Various elements of the TED parallel text corpora for the IWSLT 2014 evaluation
campaign were used as the basis for training of language models, and for
development, tuning and testing of the translation system as well as Wikipedia
based comparable corpora prepared by us. The BLEU, NIST, METEOR and TER metrics
were used to evaluate the effects of data preparations on translation results.
Our experiments included systems, which use lemma and morphological information
on Polish words. We also conducted a deep analysis of provided Polish data as
preparatory work for the automatic data correction and cleaning phase.
| 2,015 | Computation and Language |
Building Subject-aligned Comparable Corpora and Mining it for Truly
Parallel Sentence Pairs | Parallel sentences are a relatively scarce but extremely useful resource for
many applications including cross-lingual retrieval and statistical machine
translation. This research explores our methodology for mining such data from
previously obtained comparable corpora. The task is highly practical since
non-parallel multilingual data exist in far greater quantities than parallel
corpora, but parallel sentences are a much more useful resource. Here we
propose a web crawling method for building subject-aligned comparable corpora
from Wikipedia articles. We also introduce a method for extracting truly
parallel sentences that are filtered out from noisy or just comparable sentence
pairs. We describe our implementation of a specialized tool for this task as
well as training and adaption of a machine translation system that supplies our
filter with additional information about the similarity of comparable sentence
pairs.
| 2,014 | Computation and Language |
Polish -English Statistical Machine Translation of Medical Texts | This new research explores the effects of various training methods on a
Polish to English Statistical Machine Translation system for medical texts.
Various elements of the EMEA parallel text corpora from the OPUS project were
used as the basis for training of phrase tables and language models and for
development, tuning and testing of the translation system. The BLEU, NIST,
METEOR, RIBES and TER metrics have been used to evaluate the effects of various
system and data preparations on translation results. Our experiments included
systems that used POS tagging, factored phrase models, hierarchical models,
syntactic taggers, and many different alignment methods. We also conducted a
deep analysis of Polish data as preparatory work for automatic data correction
such as true casing and punctuation normalization phase.
| 2,015 | Computation and Language |
Very Deep Multilingual Convolutional Neural Networks for LVCSR | Convolutional neural networks (CNNs) are a standard component of many current
state-of-the-art Large Vocabulary Continuous Speech Recognition (LVCSR)
systems. However, CNNs in LVCSR have not kept pace with recent advances in
other domains where deeper neural networks provide superior performance. In
this paper we propose a number of architectural advances in CNNs for LVCSR.
First, we introduce a very deep convolutional network architecture with up to
14 weight layers. There are multiple convolutional layers before each pooling
layer, with small 3x3 kernels, inspired by the VGG Imagenet 2014 architecture.
Then, we introduce multilingual CNNs with multiple untied layers. Finally, we
introduce multi-scale input features aimed at exploiting more context at
negligible computational cost. We evaluate the improvements first on a Babel
task for low resource speech recognition, obtaining an absolute 5.77% WER
improvement over the baseline PLP DNN by training our CNN on the combined data
of six different languages. We then evaluate the very deep CNNs on the Hub5'00
benchmark (using the 262 hours of SWB-1 training data) achieving a word error
rate of 11.8% after cross-entropy training, a 1.4% WER improvement (10.6%
relative) over the best published CNN result so far.
| 2,016 | Computation and Language |
Enhanced Bilingual Evaluation Understudy | Our research extends the Bilingual Evaluation Understudy (BLEU) evaluation
technique for statistical machine translation to make it more adjustable and
robust. We intend to adapt it to resemble human evaluation more. We perform
experiments to evaluate the performance of our technique against the primary
existing evaluation methods. We describe and show the improvements it makes
over existing methods as well as correlation to them. When human translators
translate a text, they often use synonyms, different word orders or style, and
other similar variations. We propose an SMT evaluation technique that enhances
the BLEU metric to consider variations such as those.
| 2,015 | Computation and Language |
Real-Time Statistical Speech Translation | This research investigates the Statistical Machine Translation approaches to
translate speech in real time automatically. Such systems can be used in a
pipeline with speech recognition and synthesis software in order to produce a
real-time voice communication system between foreigners. We obtained three main
data sets from spoken proceedings that represent three different types of human
speech. TED, Europarl, and OPUS parallel text corpora were used as the basis
for training of language models, for developmental tuning and testing of the
translation system. We also conducted experiments involving part of speech
tagging, compound splitting, linear language model interpolation, TrueCasing
and morphosyntactic analysis. We evaluated the effects of variety of data
preparations on the translation results using the BLEU, NIST, METEOR and TER
metrics and tried to give answer which metric is most suitable for PL-EN
language pair.
| 2,014 | Computation and Language |
A Sentence Meaning Based Alignment Method for Parallel Text Corpora
Preparation | Text alignment is crucial to the accuracy of Machine Translation (MT)
systems, some NLP tools or any other text processing tasks requiring bilingual
data. This research proposes a language independent sentence alignment approach
based on Polish (not position-sensitive language) to English experiments. This
alignment approach was developed on the TED Talks corpus, but can be used for
any text domain or language pair. The proposed approach implements various
heuristics for sentence recognition. Some of them value synonyms and semantic
text structure analysis as a part of additional information. Minimization of
data loss was ensured. The solution is compared to other sentence alignment
implementations. Also an improvement in MT system score with text processed
with described tool is shown.
| 2,014 | Computation and Language |
Polish - English Speech Statistical Machine Translation Systems for the
IWSLT 2013 | This research explores the effects of various training settings from Polish
to English Statistical Machine Translation system for spoken language. Various
elements of the TED parallel text corpora for the IWSLT 2013 evaluation
campaign were used as the basis for training of language models, and for
development, tuning and testing of the translation system. The BLEU, NIST,
METEOR and TER metrics were used to evaluate the effects of data preparations
on translation results. Our experiments included systems, which use stems and
morphological information on Polish words. We also conducted a deep analysis of
provided Polish data as preparatory work for the automatic data correction and
cleaning phase.
| 2,013 | Computation and Language |
The "handedness" of language: Directional symmetry breaking of sign
usage in words | Language, which allows complex ideas to be communicated through symbolic
sequences, is a characteristic feature of our species and manifested in a
multitude of forms. Using large written corpora for many different languages
and scripts, we show that the occurrence probability distributions of signs at
the left and right ends of words have a distinct heterogeneous nature.
Characterizing this asymmetry using quantitative inequality measures, viz.
information entropy and the Gini index, we show that the beginning of a word is
less restrictive in sign usage than the end. This property is not simply
attributable to the use of common affixes as it is seen even when only word
roots are considered. We use the existence of this asymmetry to infer the
direction of writing in undeciphered inscriptions that agrees with the
archaeological evidence. Unlike traditional investigations of phonotactic
constraints which focus on language-specific patterns, our study reveals a
property valid across languages and writing systems. As both language and
writing are unique aspects of our species, this universal signature may reflect
an innate feature of the human cognitive phenomenon.
| 2,018 | Computation and Language |
Polish to English Statistical Machine Translation | This research explores the effects of various training settings on a Polish
to English Statistical Machine Translation system for spoken language. Various
elements of the TED, Europarl, and OPUS parallel text corpora were used as the
basis for training of language models, for development, tuning and testing of
the translation system. The BLEU, NIST, METEOR and TER metrics were used to
evaluate the effects of the data preparations on the translation results.
| 2,013 | Computation and Language |
Determination of the Internet Anonymity Influence on the Level of
Aggression and Usage of Obscene Lexis | This article deals with the analysis of the semantic content of the anonymous
Russian-speaking forum 2ch.hk, different verbal means of expressing of the
emotional state of aggression are revealed for this site, and aggression is
classified by its directions. The lexis of different Russian-and English-
speaking anonymous forums (2ch.hk and iichan.hk, 4chan.org) and public
community "MDK" of the Russian-speaking social network VK is analyzed and
compared with the Open Corpus of the Russian language (Opencorpora.org and
Brown corpus). The analysis shows that anonymity has no influence on the amount
of invective items usage. The effectiveness of moderation was shown for
anonymous forums. It was established that Russian obscene lexis was used to
express the emotional state of aggression only in 60.4% of cases for 2ch.hk.
These preliminary results show that the Russian obscene lexis on the Internet
does not have direct dependence on the emotional state of aggression.
| 2,015 | Computation and Language |
A Generative Model of Words and Relationships from Multiple Sources | Neural language models are a powerful tool to embed words into semantic
vector spaces. However, learning such models generally relies on the
availability of abundant and diverse training examples. In highly specialised
domains this requirement may not be met due to difficulties in obtaining a
large corpus, or the limited range of expression in average use. Such domains
may encode prior knowledge about entities in a knowledge base or ontology. We
propose a generative model which integrates evidence from diverse data sources,
enabling the sharing of semantic information. We achieve this by generalising
the concept of co-occurrence from distributional semantics to include other
relationships between entities or words, which we model as affine
transformations on the embedding space. We demonstrate the effectiveness of
this approach by outperforming recent models on a link prediction task and
demonstrating its ability to profit from partially or fully unobserved data
training labels. We further demonstrate the usefulness of learning from
different data sources with overlapping vocabularies.
| 2,015 | Computation and Language |
Response to Liu, Xu, and Liang (2015) and Ferrer-i-Cancho and
G\'omez-Rodr\'iguez (2015) on Dependency Length Minimization | We address recent criticisms (Liu et al., 2015; Ferrer-i-Cancho and
G\'omez-Rodr\'iguez, 2015) of our work on empirical evidence of dependency
length minimization across languages (Futrell et al., 2015). First, we
acknowledge error in failing to acknowledge Liu (2008)'s previous work on
corpora of 20 languages with similar aims. A correction will appear in PNAS.
Nevertheless, we argue that our work provides novel, strong evidence for
dependency length minimization as a universal quantitative property of
languages, beyond this previous work, because it provides baselines which focus
on word order preferences. Second, we argue that our choices of baselines were
appropriate because they control for alternative theories.
| 2,015 | Computation and Language |
Automatic Taxonomy Extraction from Query Logs with no Additional Sources
of Information | Search engine logs store detailed information on Web users interactions.
Thus, as more and more people use search engines on a daily basis, important
trails of users common knowledge are being recorded in those files. Previous
research has shown that it is possible to extract concept taxonomies from full
text documents, while other scholars have proposed methods to obtain similar
queries from query logs. We propose a mixture of both lines of research, that
is, mining query logs not to find related queries nor query hierarchies, but
actual term taxonomies that could be used to improve search engine
effectiveness and efficiency. As a result, in this study we have developed a
method that combines lexical heuristics with a supervised classification model
to successfully extract hyponymy relations from specialization search patterns
revealed from log missions, with no additional sources of information, and in a
language independent way.
| 2,015 | Computation and Language |
A Primer on Neural Network Models for Natural Language Processing | Over the past few years, neural networks have re-emerged as powerful
machine-learning models, yielding state-of-the-art results in fields such as
image recognition and speech processing. More recently, neural network models
started to be applied also to textual natural language signals, again with very
promising results. This tutorial surveys neural network models from the
perspective of natural language processing research, in an attempt to bring
natural-language researchers up to speed with the neural techniques. The
tutorial covers input encoding for natural language tasks, feed-forward
networks, convolutional networks, recurrent networks and recursive networks, as
well as the computation graph abstraction for automatic gradient computation.
| 2,015 | Computation and Language |
It is not all downhill from here: Syllable Contact Law in Persian | Syllable contact pairs crosslinguistically tend to have a falling sonority
slope a constraint which is called the Syllable Contact Law SCL In this study
the phonotactics of syllable contacts in 4202 CVCCVC words of Persian lexicon
is investigated The consonants of Persian were divided into five sonority
categories and the frequency of all possible sonority slopes is computed both
in lexicon type frequency and in corpus token frequency Since an unmarked
phonological structure has been shown to diachronically become more frequent we
expect to see the same pattern for syllable contact pairs with falling sonority
slope The correlation of sonority categories of the two consonants in a
syllable contact pair is measured using Pointwise Mutual Information
| 2,015 | Computation and Language |
P-trac Procedure: The Dispersion and Neutralization of Contrasts in
Lexicon | Cognitive acoustic cues have an important role in shaping the phonological
structure of language as a means to optimal communication. In this paper we
introduced P-trac procedure in order to track dispersion of contrasts in
different contexts in lexicon. The results of applying P-trac procedure to the
case of dispersion of contrasts in pre- consonantal contexts and in consonantal
positions of CVCC sequences in Persian provide Evidence in favor of phonetic
basis of dispersion argued by Licensing by Cue hypothesis and the Dispersion
Theory of Contrast. The P- trac procedure is proved to be very effective in
revealing the dispersion of contrasts in lexicon especially when comparing the
dispersion of contrasts in different contexts.
| 2,015 | Computation and Language |
Deep convolutional acoustic word embeddings using word-pair side
information | Recent studies have been revisiting whole words as the basic modelling unit
in speech recognition and query applications, instead of phonetic units. Such
whole-word segmental systems rely on a function that maps a variable-length
speech segment to a vector in a fixed-dimensional space; the resulting acoustic
word embeddings need to allow for accurate discrimination between different
word types, directly in the embedding space. We compare several old and new
approaches in a word discrimination task. Our best approach uses side
information in the form of known word pairs to train a Siamese convolutional
neural network (CNN): a pair of tied networks that take two speech segments as
input and produce their embeddings, trained with a hinge loss that separates
same-word pairs and different-word pairs by some margin. A word classifier CNN
performs similarly, but requires much stronger supervision. Both types of CNNs
yield large improvements over the best previously published results on the word
discrimination task.
| 2,016 | Computation and Language |
Stochastic model for phonemes uncovers an author-dependency of their
usage | We study rank-frequency relations for phonemes, the minimal units that still
relate to linguistic meaning. We show that these relations can be described by
the Dirichlet distribution, a direct analogue of the ideal-gas model in
statistical mechanics. This description allows us to demonstrate that the
rank-frequency relations for phonemes of a text do depend on its author. The
author-dependency effect is not caused by the author's vocabulary (common words
used in different texts), and is confirmed by several alternative means. This
suggests that it can be directly related to phonemes. These features contrast
to rank-frequency relations for words, which are both author and text
independent and are governed by the Zipf's law.
| 2,016 | Computation and Language |
Analyzer and generator for Pali | This work describes a system that performs morphological analysis and
generation of Pali words. The system works with regular inflectional paradigms
and a lexical database. The generator is used to build a collection of
inflected and derived words, which in turn is used by the analyzer. Generating
and storing morphological forms along with the corresponding morphological
information allows for efficient and simple look up by the analyzer. Indeed, by
looking up a word and extracting the attached morphological information, the
analyzer does not have to compute this information. As we must, however, assume
the lexical database to be incomplete, the system can also work without the
dictionary component, using a rule-based approach.
| 2,015 | Computation and Language |
Language Segmentation | Language segmentation consists in finding the boundaries where one language
ends and another language begins in a text written in more than one language.
This is important for all natural language processing tasks. The problem can be
solved by training language models on language data. However, in the case of
low- or no-resource languages, this is problematic. I therefore investigate
whether unsupervised methods perform better than supervised methods when it is
difficult or impossible to train supervised approaches. A special focus is
given to difficult texts, i.e. texts that are rather short (one sentence),
containing abbreviations, low-resource languages and non-standard language. I
compare three approaches: supervised n-gram language models, unsupervised
clustering and weakly supervised n-gram language model induction. I devised the
weakly supervised approach in order to deal with difficult text specifically.
In order to test the approach, I compiled a small corpus of different text
types, ranging from one-sentence texts to texts of about 300 words. The weakly
supervised language model induction approach works well on short and difficult
texts, outperforming the clustering algorithm and reaching scores in the
vicinity of the supervised approach. The results look promising, but there is
room for improvement and a more thorough investigation should be undertaken.
| 2,015 | Computation and Language |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.