Datasets:

arbml

/

masader

like 7

Twifil

nan

https://github.com/kinmokusu/oea_algd

unknown

ar

ar-DZ: (Arabic (Algeria))

social media

text

crawling and annotation(other)

An Algerian dialect dataset annotated for both sentiment (9,000 tweets), emotion (about 5,000 tweets) and extra-linguistic information including author profiling (age and gender)

14,000

sentences

Low

nan

nan

An Algerian Corpus and an Annotation Platform for Opinion and Emotion Analysis

https://aclanthology.org/2020.lrec-1.151.pdf

Arab-Latn

No

GitHub

Free

nan

No

sentiment analysis, Emotion detection

LREC

14.0

conference

Language Resources and Evaluation Conference

Leila Moudjari, Karima Akli-Astouati, Farah Benamara

nan

In this paper, we address the lack of resources for opinion and emotion analysis related to North African dialects, targeting Algerian dialect. We present TWIFIL (TWItter proFILing) a collaborative annotation platform for crowdsourcing annotation of tweets at different levels of granularity. The plateform allowed the creation of the largest Algerian dialect dataset annotated for both sentiment (9,000 tweets), emotion (about 5,000 tweets) and extra-linguistic information including author profiling (age and gender). The annotation resulted also in the creation of the largest Algerien dialect subjectivity lexicon of about 9,000 entries which can constitute a valuable resources for the development of future NLP applications for Algerian dialect. To test the validity of the dataset, a set of deep learning experiments were conducted to classify a given tweet as positive, negative or neutral. We discuss our results and provide an error analysis to better identify classification errors.

Abderrahmane Issam

MADAR Lexicon

nan

https://docs.google.com/forms/d/e/1FAIpQLSe2LHYmHsxdkHPYHgcZDz25dTNbnygPkmClIaLd_fwud-XnTQ/viewform

custom

ar

mixed

other

text

manual curation

The MADAR Lexicon is a collection of 1,042 concepts expressed in 25 city dialects totaling 47K entries (with an average of 45 words per concept, or about 2 words per dialect). Concepts were selected from the BTEC Parallel corpora. The lexicon is centered around concept keys, which are triplets of English, French, and Modern Standard Arabic (MSA), and annotators had to provide words that overlap in word sense with all three languages. Each dialectal word is presented in its CODA orthography and its CAPHI phonology (Bouamor et al., 2018; Habash et al., 2018). The MADAR Lexicon was created as part of the Multi-Arabic Dialect Applications and Resources Project (funded by NPRP 7-290- 1-047 from the Qatar National Research Fund (a member of the Qatar Foundation). Website: http://madar.camel-lab.com

47,000

tokens

Low

NYU Abu Dhabi

nan

The MADAR Arabic Dialect Corpus and Lexicon.

http://www.lrec-conf.org/proceedings/lrec2018/pdf/351.pdf

Arab

Yes

CAMeL Resources

Free

nan

No

dialect identification, transliteration

LREC

127.0

conference

The International Conference on Language Resources and Evaluation

Bouamor, Houda, Nizar Habash, Mohammad Salameh, Wajdi Zaghouani, Owen Rambow, Dana Abdulrahim, Ossama Obeid, Salam Khalifa, Fadhl Eryani, Alexander Erdmann and Kemal Oflazer.

Carnegie Mellon University in Qatar, Qatar; Hamad Bin Khalifa University, Qatar; New York University Abu Dhabi, UAE; Columbia University, USA, University of Bahrain; Bahrain

In this paper, we present two resources that were created as part of the Multi Arabic Dialect Applications and Resources (MADAR) project. The first is a large parallel corpus of 25 Arabic city dialects in the travel domain. The second is a lexicon of 1,045 concepts with an average of 45 words from 25 cities per concept. These resources are the first of their kind in terms of the breadth of their coverage and the fine location granularity. The focus on cities, as opposed to regions in studying Arabic dialects, opens new avenues to many areas of research from dialectology to dialect identification and machine translation.

Fadhl Al-Eryani

xSID - (X) Slot and Intent Detection

nan

https://bitbucket.org/robvanderg/xsid/src/master/

CC BY-SA 4.0

multilingual

ar-MSA: (Arabic (Modern Standard Arabic))

other

text

human translation

An evaluation dataset of intent classification and slot detection

800

tokens

Low

nan

nan

From Masked Language Modeling to Translation: Non-English Auxiliary Tasks Improve Zero-shot Spoken Language Understanding

https://aclanthology.org/2021.naacl-main.197.pdf

Arab

No

other

Free

nan

Yes

intent classification, slot detection

ACL

nan

conference

Associations of computation linguistics

nan

nan

The lack of publicly available evaluation data for low-resource languages limits progress in Spoken Language Understanding (SLU). As key tasks like intent classification and slot filling require abundant training data, it is desirable to reuse existing data in high-resource languages to develop models for low-resource scenarios. We introduce XSID, a new benchmark for cross-lingual (X) Slot and Intent Detection in 13 languages from 6 language families, including a very low-resource dialect. To tackle the challenge, we propose a joint learning approach, with English SLU training data and non-English auxiliary tasks from raw text, syntax and translation for transfer. We study two setups which differ by type and language coverage of the pre-trained embeddings. Our results show that jointly learning the main tasks with masked language modeling is effective for slots, while machine translation transfer works best for intent classification

Amr Keleg

Annotated Shami Corpus

nan

https://github.com/christios/annotated-shami-corpus

CC BY 4.0

ar

ar-LB: (Arabic (Lebanon))

social media

text

crawling and annotation(other)

Subsection of the Lebanese portion of the Shami Corpus annotated for spelling standardization (CODA), morphological segmentation and tagging, and spontaneous orthography taxonomy tagging.

10,000

tokens

Medium

nan

Shami Corpus

Orthography Standardization in Arabic Dialects

https://dspace.cuni.cz/handle/20.500.11956/147949

Arab

Yes

GitHub

Free

nan

No

part of speech tagging, morphological analysis, error class taxonomy tagging, CODA

nan

nan

nan

nan

Christian Khairallah

Charles University in Prague, Saarland University

Spontaneous orthography in Arabic dialects poses one of the biggest obstacles in the way of Dialectal Arabic NLP applications. As the Arab world enjoys a wide array of these widely spoken and recently written, non-standard, low-resource varieties, this thesis presents a detailed account of this relatively overlooked phenomenon. It sets out to show that continuously creating additional noise-free, manually standardized corpora of Dialectal Arabic does not free us from the shackles of non-standard (spontaneous) orthography. Because real-world data will most often come in a noisy format, it also investigates ways to ease the amount of noise in textual data. As a proof of concept, we restrict ourselves to one of the dialectal varieties, namely, Lebanese Arabic. It also strives to gain a better understanding of the nature of the noise and its distribution. All of this is done by leveraging various spelling correction and morphological tagging neural architectures in a multi-task setting, and by annotating a Lebanese Arabic corpus for spontaneous orthography standardization, and morphological segmentation and tagging, among other features. Additionally, a detailed taxonomy of spelling inconsistencies for Lebanese Arabic is presented and is used to tag the corpus. This constitutes the first attempt in Dialectal Arabic research to try and categorize spontaneous orthography in a detailed manner.

Christian Khairallah

Maknuune: A Large Open Palestinian Arabic Lexicon

nan

http://www.palestine-lexicon.org/

CC BY-SA 4.0

ar

ar-PS: (Arabic (Palestine))

other

spoken

manual curation

Palestinian Arabic lexicon collected through manual curation and field surveys.

36,302

tokens

Low

Shahd Dibas and NYU Abu Dhabi

nan

Maknuune: A Large Open Palestinian Arabic Lexicon

https://arxiv.org/pdf/2210.12985.pdf

Arab

No

Gdrive

Free

nan

No

Lexicon (linguistic analysis)

WANLP

nan

workshop

Arabic Natural Language Processing Workshop

Shahd Dibas, Christian Khairallah, Nizar Habash, Omar Fayez Sadi, Tariq Sairafy, Karmel Sarabta, Abrar Ardah

NYUAD, University of Oxford, UNRWA

We present Maknuune, a large open lexicon for the Palestinian Arabic dialect. Maknuune has over 36K entries from 17K lemmas, and 3.7K roots. All entries include diacritized Arabic orthography, phonological transcription and English glosses. Some entries are enriched with additional information such as broken plurals and templatic feminine forms, associated phrases and collocations, Standard Arabic glosses, and examples or notes on grammar, usage, or location of collected entry.

Christian Khairallah

Baladi Lebanese dialect corpora

nan

https://portal.sina.birzeit.edu/curras

CC BY-NC-SA 4.0

ar

ar-LB: (Arabic (Lebanon))

other

text

crawling

The corpus consists of about 9.6K words/tokens collected from Facebook, blog posts and traditional poems. The corpus was annotated as an extension to Curras and following the same annotation methodology to form a Levantine Corpus.

10,000

tokens

Low

Birzeit University

nan

Curras + Baladi: Towards a Levantine Corpus

https://arxiv.org/pdf/2212.06468.pdf

Arab

Yes

Dropbox

Upon-Request

nan

No

machine translation, speech recognition, dialect identification, named entity recognition, part of speech tagging, language identification, morphological analysis

LERC

nan

conference

LERC

Karim El Haff, Mustafa Jarrar, Tymaa Hammouda, Fadi Zaraket

nan

The processing of the Arabic language is a complex field of research. This is due to many factors, including the complex and rich morphology of Arabic, its high degree of ambiguity, and the presence of several regional varieties that need to be processed while taking into account their unique characteristics. When its dialects are taken into account, this language pushes the limits of NLP to find solutions to problems posed by its inherent nature. It is a diglossic language; the standard language is used in formal settings and in education and is quite different from the vernacular languages spoken in the different regions and influenced by older languages that were historically spoken in those regions. This should encourage NLP specialists to create dialect-specific corpora such as the Palestinian morphologically annotated Curras corpus of Birzeit University. In this work, we present the Lebanese Corpus Baladi that consists of around 9.6K morphologically annotated tokens. Since Lebanese and Palestinian dialects are part of the same Levantine dialectal continuum, and thus highly mutually intelligible, our proposed corpus was constructed to be used to (1) enrich Curras and transform it into a more general Levantine corpus and (2) improve Curras by solving detected errors.

Mustafa Jarrar

Mawqif

https://huggingface.co/datasets/NoraAlt/Mawqif_Stance-Detection

https://github.com/NoraAlt/Mawqif-Arabic-Stance

unknown

ar

mixed

social media

text

crawling and annotation(other)

Mawqif is the first Arabic dataset that can be used for target-specific stance detection. This is a multi-label dataset where each data point is annotated for stance, sentiment, and sarcasm.

4,121

sentences

Medium

nan

nan

Mawqif: A Multi-label Arabic Dataset for Target-specific Stance Detection

https://aclanthology.org/2022.wanlp-1.16/

Arab

No

GitHub

Free

nan

Yes

sentiment analysis, topic classification, irony detection, stance detection

WANLP

nan

workshop

Arabic Natural Language Processing Workshop

nan

nan

Social media platforms are becoming inherent parts of people’s daily life to express opinions and stances toward topics of varying polarities. Stance detection determines the viewpoint expressed in a text toward a target. While communication on social media (e.g., Twitter) takes place in more than 40 languages, the majority of stance detection research has been focused on English. Although some efforts have recently been made to develop stance detection datasets in other languages, no similar efforts seem to have considered the Arabic language. In this paper, we present Mawqif, the first Arabic dataset for target-specific stance detection, composed of 4,121 tweets annotated with stance, sentiment, and sarcasm polarities. Mawqif, as a multi-label dataset, can provide more opportunities for studying the interaction between different opinion dimensions and evaluating a multi-task model. We provide a detailed description of the dataset, present an analysis of the produced annotation, and evaluate four BERT-based models on it. Our best model achieves a macro-F1 of 78.89%, which shows that there is ample room for improvement on this challenging task. We publicly release our dataset, the annotation guidelines, and the code of the experiments.

Nora Saleh Alturayeif

Tatoeba

nan

https://tatoeba.org/en/downloads

CC BY 2.0

multilingual

mixed

other

text

manual curation

A crowd-sourced dataset of parallel sentences.

nan

sentences

nan

nan

nan

nan

nan

Arab

No

other

Free

nan

No

machine translation, language modelling, dialect identification, language identification

nan

nan

nan

nan

nan

nan

nan

Amr Keleg

QA4MRE

https://huggingface.co/datasets/qa4mre

http://nlp.uned.es/clef-qa/repository/qa4mre.php

unknown

multilingual

ar-MSA: (Arabic (Modern Standard Arabic))

other

text

crawling and annotation(other)

QA4MRE dataset was created for the CLEF 2011/2012/2013 shared tasks to promote research in question answering and reading comprehension. The dataset contains a supporting passage and a set of questions corresponding to the passage. Multiple options for answers are provided for each question, of which only one is correct. The training and test datasets are available for the main track. Additional gold standard documents are available for two pilot studies: one on alzheimers data, and the other on entrance exams data.

160

documents

Low

nan

nan

QA4MRE 2011-2013: Overview of Question Answering for Machine Reading Evaluation

https://link.springer.com/chapter/10.1007/978-3-642-40802-1_29

Arab

No

other

Free

nan

No

multiple choice

CLEF

nan

conference

Conference and Labs of the Evaluation Forum

Anselmo Peñas, Eduard Hovy, Pamela Forner, Álvaro Rodrigo, Richard Sutcliffe & Roser Morante

nan

This paper describes the methodology for testing the performance of Machine Reading systems through Question Answering and Reading Comprehension Tests. This was the attempt of the QA4MRE challenge which was run as a Lab at CLEF 2011–2013. The traditional QA task was replaced by a new Machine Reading task, whose intention was to ask questions that required a deep knowledge of individual short texts and in which systems were required to choose one answer, by analysing the corresponding test document in conjunction with background text collections provided by the organization. Four different tasks have been organized during these years: Main Task, Processing Modality and Negation for Machine Reading, Machine Reading of Biomedical Texts about Alzheimer’s disease, and Entrance Exams. This paper describes their motivation, their goals, their methodology for preparing the data sets, their background collections, their metrics used for the evaluation, and the lessons learned along these three years.

Zaid Alyafeai