Datasets:

QCRI
/

LlamaLens-Arabic

	---
	license: cc-by-nc-sa-4.0
	task_categories:
	- text-classification
	language:
	- ar
	tags:
	- Social Media
	- News Media
	- Sentiment
	- Stance
	- Emotion
	pretty_name: 'LlamaLens: Specialized Multilingual LLM for Analyzing News and Social Media Content -- Arabic'
	size_categories:
	- 10K<n<100K
	dataset_info:
	- config_name: SANADAkhbarona-news-categorization
	splits:
	- name: train
	num_examples: 62210
	- name: dev
	num_examples: 7824
	- name: test
	num_examples: 7824
	- config_name: CT22Harmful
	splits:
	- name: train
	num_examples: 2484
	- name: dev
	num_examples: 1076
	- name: test
	num_examples: 1201
	- config_name: Mawqif-Arabic-Stance-main
	splits:
	- name: train
	num_examples: 3162
	- name: dev
	num_examples: 950
	- name: test
	num_examples: 560
	- config_name: CT22Claim
	splits:
	- name: train
	num_examples: 3513
	- name: dev
	num_examples: 339
	- name: test
	num_examples: 1248
	- config_name: annotated-hatetweets-4-classes
	splits:
	- name: train
	num_examples: 210525
	- name: dev
	num_examples: 90543
	- name: test
	num_examples: 100564
	- config_name: ar_reviews_100k
	splits:
	- name: train
	num_examples: 69998
	- name: dev
	num_examples: 10000
	- name: test
	num_examples: 20000
	- config_name: Arafacts
	splits:
	- name: train
	num_examples: 4354
	- name: dev
	num_examples: 623
	- name: test
	num_examples: 1245
	- config_name: OSACT4SubtaskA
	splits:
	- name: train
	num_examples: 4780
	- name: dev
	num_examples: 2047
	- name: test
	num_examples: 1827
	- config_name: SANADAlArabiya-news-categorization
	splits:
	- name: train
	num_examples: 56967
	- name: dev
	num_examples: 7120
	- name: test
	num_examples: 7123
	- config_name: ArPro
	splits:
	- name: train
	num_examples: 6002
	- name: dev
	num_examples: 672
	- name: test
	num_examples: 1326
	- config_name: xlsum
	splits:
	- name: train
	num_examples: 37425
	- name: dev
	num_examples: 4689
	- name: test
	num_examples: 4689
	- config_name: ArSarcasm-v2
	splits:
	- name: train
	num_examples: 8749
	- name: dev
	num_examples: 3761
	- name: test
	num_examples: 2996
	- config_name: COVID19Factuality
	splits:
	- name: train
	num_examples: 3513
	- name: dev
	num_examples: 339
	- name: test
	num_examples: 988
	- config_name: Emotional-Tone
	splits:
	- name: train
	num_examples: 7024
	- name: dev
	num_examples: 1005
	- name: test
	num_examples: 2009
	- config_name: ans-claim
	splits:
	- name: train
	num_examples: 3185
	- name: dev
	num_examples: 906
	- name: test
	num_examples: 456
	- config_name: ArCyc_OFF
	splits:
	- name: train
	num_examples: 3138
	- name: dev
	num_examples: 450
	- name: test
	num_examples: 900
	- config_name: CT24_checkworthy
	splits:
	- name: train
	num_examples: 7333
	- name: dev
	num_examples: 1093
	- name: test
	num_examples: 610
	- config_name: stance
	splits:
	- name: train
	num_examples: 2652
	- name: dev
	num_examples: 755
	- name: test
	num_examples: 379
	- config_name: NewsHeadline
	splits:
	- name: train
	num_examples: 939
	- name: dev
	num_examples: 160
	- name: test
	num_examples: 323
	- config_name: NewsCredibilityDataset
	splits:
	- name: train
	num_examples: 8671
	- name: dev
	num_examples: 1426
	- name: test
	num_examples: 2730
	- config_name: UltimateDataset
	splits:
	- name: train
	num_examples: 133036
	- name: dev
	num_examples: 19269
	- name: test
	num_examples: 38456
	- config_name: ThatiAR
	splits:
	- name: train
	num_examples: 2446
	- name: dev
	num_examples: 467
	- name: test
	num_examples: 748
	- config_name: ArSAS
	splits:
	- name: train
	num_examples: 13883
	- name: dev
	num_examples: 1987
	- name: test
	num_examples: 3976
	- config_name: CT22Attentionworthy
	splits:
	- name: train
	num_examples: 2479
	- name: dev
	num_examples: 1071
	- name: test
	num_examples: 1186
	- config_name: ASND
	splits:
	- name: train
	num_examples: 74496
	- name: dev
	num_examples: 11136
	- name: test
	num_examples: 21942
	- config_name: OSACT4SubtaskB
	splits:
	- name: train
	num_examples: 4778
	- name: dev
	num_examples: 2048
	- name: test
	num_examples: 1827
	- config_name: ArCyc_CB
	splits:
	- name: train
	num_examples: 3145
	- name: dev
	num_examples: 451
	- name: test
	num_examples: 900
	- config_name: SANADAlkhaleej-news-categorization
	splits:
	- name: train
	num_examples: 36391
	- name: dev
	num_examples: 4550
	- name: test
	num_examples: 4550
	configs:
	- config_name: SANADAkhbarona-news-categorization
	data_files:
	- split: test
	path: SANADAkhbarona-news-categorization/test.json
	- split: dev
	path: SANADAkhbarona-news-categorization/dev.json
	- split: train
	path: SANADAkhbarona-news-categorization/train.json
	- config_name: CT22Harmful
	data_files:
	- split: test
	path: CT22Harmful/test.json
	- split: dev
	path: CT22Harmful/dev.json
	- split: train
	path: CT22Harmful/train.json
	- config_name: Mawqif-Arabic-Stance-main
	data_files:
	- split: test
	path: Mawqif-Arabic-Stance-main/test.json
	- split: dev
	path: Mawqif-Arabic-Stance-main/dev.json
	- split: train
	path: Mawqif-Arabic-Stance-main/train.json
	- config_name: CT22Claim
	data_files:
	- split: test
	path: CT22Claim/test.json
	- split: dev
	path: CT22Claim/dev.json
	- split: train
	path: CT22Claim/train.json
	- config_name: annotated-hatetweets-4-classes
	data_files:
	- split: test
	path: annotated-hatetweets-4-classes/test.json
	- split: dev
	path: annotated-hatetweets-4-classes/dev.json
	- split: train
	path: annotated-hatetweets-4-classes/train.json
	- config_name: ar_reviews_100k
	data_files:
	- split: test
	path: ar_reviews_100k/test.json
	- split: dev
	path: ar_reviews_100k/dev.json
	- split: train
	path: ar_reviews_100k/train.json
	- config_name: Arafacts
	data_files:
	- split: test
	path: Arafacts/test.json
	- split: dev
	path: Arafacts/dev.json
	- split: train
	path: Arafacts/train.json
	- config_name: OSACT4SubtaskA
	data_files:
	- split: test
	path: OSACT4SubtaskA/test.json
	- split: dev
	path: OSACT4SubtaskA/dev.json
	- split: train
	path: OSACT4SubtaskA/train.json
	- config_name: SANADAlArabiya-news-categorization
	data_files:
	- split: test
	path: SANADAlArabiya-news-categorization/test.json
	- split: dev
	path: SANADAlArabiya-news-categorization/dev.json
	- split: train
	path: SANADAlArabiya-news-categorization/train.json
	- config_name: ArPro
	data_files:
	- split: test
	path: ArPro/test.json
	- split: dev
	path: ArPro/dev.json
	- split: train
	path: ArPro/train.json
	- config_name: xlsum
	data_files:
	- split: test
	path: xlsum/test.json
	- split: dev
	path: xlsum/dev.json
	- split: train
	path: xlsum/train.json
	- config_name: ArSarcasm-v2
	data_files:
	- split: test
	path: ArSarcasm-v2/test.json
	- split: dev
	path: ArSarcasm-v2/dev.json
	- split: train
	path: ArSarcasm-v2/train.json
	- config_name: COVID19Factuality
	data_files:
	- split: test
	path: COVID19Factuality/test.json
	- split: dev
	path: COVID19Factuality/dev.json
	- split: train
	path: COVID19Factuality/train.json
	- config_name: Emotional-Tone
	data_files:
	- split: test
	path: Emotional-Tone/test.json
	- split: dev
	path: Emotional-Tone/dev.json
	- split: train
	path: Emotional-Tone/train.json
	- config_name: ans-claim
	data_files:
	- split: test
	path: ans-claim/test.json
	- split: dev
	path: ans-claim/dev.json
	- split: train
	path: ans-claim/train.json
	- config_name: ArCyc_OFF
	data_files:
	- split: test
	path: ArCyc_OFF/test.json
	- split: dev
	path: ArCyc_OFF/dev.json
	- split: train
	path: ArCyc_OFF/train.json
	- config_name: CT24_checkworthy
	data_files:
	- split: test
	path: CT24_checkworthy/test.json
	- split: dev
	path: CT24_checkworthy/dev.json
	- split: train
	path: CT24_checkworthy/train.json
	- config_name: stance
	data_files:
	- split: test
	path: stance/test.json
	- split: dev
	path: stance/dev.json
	- split: train
	path: stance/train.json
	- config_name: NewsHeadline
	data_files:
	- split: test
	path: NewsHeadline/test.json
	- split: dev
	path: NewsHeadline/dev.json
	- split: train
	path: NewsHeadline/train.json
	- config_name: NewsCredibilityDataset
	data_files:
	- split: test
	path: NewsCredibilityDataset/test.json
	- split: dev
	path: NewsCredibilityDataset/dev.json
	- split: train
	path: NewsCredibilityDataset/train.json
	- config_name: UltimateDataset
	data_files:
	- split: test
	path: UltimateDataset/test.json
	- split: dev
	path: UltimateDataset/dev.json
	- split: train
	path: UltimateDataset/train.json
	- config_name: ThatiAR
	data_files:
	- split: test
	path: ThatiAR/test.json
	- split: dev
	path: ThatiAR/dev.json
	- split: train
	path: ThatiAR/train.json
	- config_name: ArSAS
	data_files:
	- split: test
	path: ArSAS/test.json
	- split: dev
	path: ArSAS/dev.json
	- split: train
	path: ArSAS/train.json
	- config_name: CT22Attentionworthy
	data_files:
	- split: test
	path: CT22Attentionworthy/test.json
	- split: dev
	path: CT22Attentionworthy/dev.json
	- split: train
	path: CT22Attentionworthy/train.json
	- config_name: ASND
	data_files:
	- split: test
	path: ASND/test.json
	- split: dev
	path: ASND/dev.json
	- split: train
	path: ASND/train.json
	- config_name: OSACT4SubtaskB
	data_files:
	- split: test
	path: OSACT4SubtaskB/test.json
	- split: dev
	path: OSACT4SubtaskB/dev.json
	- split: train
	path: OSACT4SubtaskB/train.json
	- config_name: ArCyc_CB
	data_files:
	- split: test
	path: ArCyc_CB/test.json
	- split: dev
	path: ArCyc_CB/dev.json
	- split: train
	path: ArCyc_CB/train.json
	- config_name: SANADAlkhaleej-news-categorization
	data_files:
	- split: test
	path: SANADAlkhaleej-news-categorization/test.json
	- split: dev
	path: SANADAlkhaleej-news-categorization/dev.json
	- split: train
	path: SANADAlkhaleej-news-categorization/train.json
	---

	# LlamaLens: Specialized Multilingual LLM Dataset

	## Overview
	LlamaLens is a specialized multilingual LLM designed for analyzing news and social media content. It focuses on 18 NLP tasks, leveraging 52 datasets across Arabic, English, and Hindi.


	<p align="center"> <img src="./capablities_tasks_datasets.png" style="width: 40%;" id="title-icon"> </p>

	## LlamaLens
	This repo includes scripts needed to run our full pipeline, including data preprocessing and sampling, instruction dataset creation, model fine-tuning, inference and evaluation.

	### Features
	- Multilingual support (Arabic, English, Hindi)
	- 18 NLP tasks with 52 datasets
	- Optimized for news and social media content analysis

	## 📂 Dataset Overview

	### Arabic Datasets

	\| Task \| Dataset \| # Labels \| # Train \| # Test \| # Dev \|
	\|---------------------------\|------------------------------\|--------------\|-------------\|------------\|-----------\|
	\| Attentionworthiness \| CT22Attentionworthy \| 9 \| 2,470 \| 1,186 \| 1,071 \|
	\| Checkworthiness \| CT24_T1 \| 2 \| 22,403 \| 500 \| 1,093 \|
	\| Claim \| CT22Claim \| 2 \| 3,513 \| 1,248 \| 339 \|
	\| Cyberbullying \| ArCyc_CB \| 2 \| 3,145 \| 900 \| 451 \|
	\| Emotion \| Emotional-Tone \| 8 \| 7,024 \| 2,009 \| 1,005 \|
	\| Emotion \| NewsHeadline \| 7 \| 939 \| 323 \| 160 \|
	\| Factuality \| Arafacts \| 5 \| 4,354 \| 1,245 \| 623 \|
	\| Factuality \| COVID19Factuality \| 2 \| 3,513 \| 988 \| 339 \|
	\| Harmful \| CT22Harmful \| 2 \| 2,484 \| 1,201 \| 1,076 \|
	\| Hate Speech \| annotated-hatetweets-4-classes \| 4 \| 210,526 \| 100,565 \| 90,544 \|
	\| Hate Speech \| OSACT4SubtaskB \| 2 \| 4,778 \| 1,827 \| 2,048 \|
	\| News Genre Categorization \| ASND \| 10 \| 74,496 \| 21,942 \| 11,136 \|
	\| News Genre Categorization \| SANADAkhbarona \| 7 \| 62,210 \| 7,824 \| 7,824 \|
	\| News Genre Categorization \| SANADAlArabiya \| 6 \| 56,967 \| 7,123 \| 7,120 \|
	\| News Genre Categorization \| SANADAlkhaleej \| 7 \| 36,391 \| 4,550 \| 4,550 \|
	\| News Genre Categorization \| UltimateDataset \| 10 \| 133,036 \| 38,456 \| 19,269 \|
	\| News Credibility \| NewsCredibilityDataset \| 2 \| 8,671 \| 2,730 \| 1,426 \|
	\| Summarization \| xlsum \| -- \| 37,425 \| 4,689 \| 4,689 \|
	\| Offensive Language \| ArCyc_OFF \| 2 \| 3,138 \| 900 \| 450 \|
	\| Offensive Language \| OSACT4SubtaskA \| 2 \| 4,780 \| 1,827 \| 2,047 \|
	\| Propaganda \| ArPro \| 2 \| 6,002 \| 1,326 \| 672 \|
	\| Sarcasm \| ArSarcasm-v2 \| 2 \| 8,749 \| 2,996 \| 3,761 \|
	\| Sentiment \| ar_reviews_100k \| 3 \| 69,998 \| 20,000 \| 10,000 \|
	\| Sentiment \| ArSAS \| 4 \| 13,883 \| 3,976 \| 1,987 \|
	\| Stance \| Mawqif-Arabic-Stance-main \| 2 \| 3,162 \| 560 \| 950 \|
	\| Stance \| stance \| 3 \| 2,652 \| 379 \| 755 \|
	\| Subjectivity \| ThatiAR \| 2 \| 2,446 \| 748 \| 467 \|


	## Results

	Below, we present the performance of L-Lens: LlamaLens , where "Eng" refers to the English-instructed model and "Native" refers to the model trained with native language instructions. The results are compared against the SOTA (where available) and the Base: Llama-Instruct 3.1 baseline. The Δ (Delta) column indicates the difference between LlamaLens and the SOTA performance, calculated as (LlamaLens – SOTA).


	---

	\| Task \| Dataset \| Metric \| SOTA \| Base \| L-Lens-Eng \| L-Lens-Native \| Δ (L-Lens (Eng) - SOTA) \|
	\|:----------------------------------:\|:--------------------------------------------:\|:----------:\|:--------:\|:---------------------:\|:---------------------:\|:--------------------:\|:------------------------:\|
	\| Attentionworthiness Detection \| CT22Attentionworthy \| W-F1 \| 0.412 \| 0.158 \| 0.425 \| 0.454 \| 0.013 \|
	\| Checkworthiness Detection \| CT24_checkworthy \| F1_Pos \| 0.569 \| 0.610 \| 0.502 \| 0.509 \| -0.067 \|
	\| Claim Detection \| CT22Claim \| Acc \| 0.703 \| 0.581 \| 0.734 \| 0.756 \| 0.031 \|
	\| Cyberbullying Detection \| ArCyc_CB \| Acc \| 0.863 \| 0.766 \| 0.870 \| 0.833 \| 0.007 \|
	\| Emotion Detection \| Emotional-Tone \| W-F1 \| 0.658 \| 0.358 \| 0.705 \| 0.736 \| 0.047 \|
	\| Emotion Detection \| NewsHeadline \| Acc \| 1.000 \| 0.406 \| 0.480 \| 0.458 \| -0.520 \|
	\| Factuality \| Arafacts \| Mi-F1 \| 0.850 \| 0.210 \| 0.771 \| 0.738 \| -0.079 \|
	\| Factuality \| COVID19Factuality \| W-F1 \| 0.831 \| 0.492 \| 0.800 \| 0.840 \| -0.031 \|
	\| Harmfulness Detection \| CT22Harmful \| F1_Pos \| 0.557 \| 0.507 \| 0.523 \| 0.535 \| -0.034 \|
	\| Hate Speech Detection \| annotated-hatetweets-4-classes \| W-F1 \| 0.630 \| 0.257 \| 0.526 \| 0.517 \| -0.104 \|
	\| Hate Speech Detection \| OSACT4SubtaskB \| Mi-F1 \| 0.950 \| 0.819 \| 0.955 \| 0.955 \| 0.005 \|
	\| News Categorization \| ASND \| Ma-F1 \| 0.770 \| 0.587 \| 0.919 \| 0.929 \| 0.149 \|
	\| News Categorization \| SANADAkhbarona-news-categorization \| Acc \| 0.940 \| 0.784 \| 0.954 \| 0.953 \| 0.014 \|
	\| News Categorization \| SANADAlArabiya-news-categorization \| Acc \| 0.974 \| 0.893 \| 0.987 \| 0.985 \| 0.013 \|
	\| News Categorization \| SANADAlkhaleej-news-categorization \| Acc \| 0.986 \| 0.865 \| 0.984 \| 0.982 \| -0.002 \|
	\| News Categorization \| UltimateDataset \| Ma-F1 \| 0.970 \| 0.376 \| 0.865 \| 0.880 \| -0.105 \|
	\| News Credibility \| NewsCredibilityDataset \| Acc \| 0.899 \| 0.455 \| 0.935 \| 0.933 \| 0.036 \|
	\| News Summarization \| xlsum \| R-2 \| 0.137 \| 0.034 \| 0.129 \| 0.130 \| -0.009 \|
	\| Offensive Language Detection \| ArCyc_OFF \| Ma-F1 \| 0.878 \| 0.489 \| 0.877 \| 0.879 \| -0.001 \|
	\| Offensive Language Detection \| OSACT4SubtaskA \| Ma-F1 \| 0.905 \| 0.782 \| 0.896 \| 0.882 \| -0.009 \|
	\| Propaganda Detection \| ArPro \| Mi-F1 \| 0.767 \| 0.597 \| 0.747 \| 0.731 \| -0.020 \|
	\| Sarcasm Detection \| ArSarcasm-v2 \| F1_Pos \| 0.584 \| 0.477 \| 0.520 \| 0.542 \| -0.064 \|
	\| Sentiment Classification \| ar_reviews_100k \| F1_Pos \| -- \| 0.681 \| 0.785 \| 0.779 \| -- \|
	\| Sentiment Classification \| ArSAS \| Acc \| 0.920 \| 0.603 \| 0.800 \| 0.804 \| -0.120 \|
	\| Stance Detection \| stance \| Ma-F1 \| 0.767 \| 0.608 \| 0.926 \| 0.881 \| 0.159 \|
	\| Stance Detection \| Mawqif-Arabic-Stance-main \| Ma-F1 \| 0.789 \| 0.764 \| 0.853 \| 0.826 \| 0.065 \|
	\| Subjectivity Detection \| ThatiAR \| f1_pos \| 0.800 \| 0.562 \| 0.441 \| 0.383 \| -0.359 \|

	---

	## File Format

	Each JSONL file in the dataset follows a structured format with the following fields:

	- `id`: Unique identifier for each data entry.
	- `original_id`: Identifier from the original dataset, if available.
	- `input`: The original text that needs to be analyzed.
	- `output`: The label assigned to the text after analysis.
	- `dataset`: Name of the dataset the entry belongs.
	- `task`: The specific task type.
	- `lang`: The language of the input text.
	- `instructions`: A brief set of instructions describing how the text should be labeled.


	Example entry in JSONL file:

	```
	{
	"id": "c64503bb-9253-4f58-aef8-9b244c088b15",
	"original_id": "1,722,643,241,323,950,300",
	"input": "يريدون توريط السلطة الفلسطينية في الضفة ودق آخر مسمار في نعش ما تبقى من هويتنا الفلسطينية، كما تم توريط غزة. يريدون إعلان كفاح مسلح من طرف الأجهزة الأمنية الفلسطينية علناً! لكن ما يعلمونه وما يرونه ولا يريدون التحدث به، أن أبناء الأجهزة الأمنية في النهار يكونون عسكريين... https://t.co/qF2Fjh24hV https://t.co/1UicLkDd52",
	"output": "checkworthy",
	"dataset": "Checkworthiness",
	"task": "Checkworthiness",
	"lang": "ar",
	"instructions": "Identify if the given factual claim is 'checkworthy' or 'not-checkworthy'. Return only the label without any explanation, justification, or additional text."
	}


	```
	## Model
	[LlamaLens on Hugging Face](https://huggingface.co/QCRI/LlamaLens)

	## Replication Scripts
	[LlamaLens GitHub Repository](https://github.com/firojalam/LlamaLens)

	## 📢 Citation

	If you use this dataset, please cite our [paper](https://arxiv.org/pdf/2410.15308):

	```
	@article{kmainasi2024llamalensspecializedmultilingualllm,
	title={LlamaLens: Specialized Multilingual LLM for Analyzing News and Social Media Content},
	author={Mohamed Bayan Kmainasi and Ali Ezzat Shahroor and Maram Hasanain and Sahinur Rahman Laskar and Naeemul Hassan and Firoj Alam},
	year={2024},
	journal={arXiv preprint arXiv:2410.15308},
	volume={},
	number={},
	pages={},
	url={https://arxiv.org/abs/2410.15308},
	eprint={2410.15308},
	archivePrefix={arXiv},
	primaryClass={cs.CL}
	}
	```