Freedom of speech on online social media platforms, often comes with the cost of hate speech production. Hate speech can be very harmful to the peace and development of societies as they bring about conflict and encourage crime. To regulate the hate speech content, moderators and annotators are employed. In our research, we look at the effects of prolonged exposure to hate speech on the mental and physical health of these annotators, as well as researchers with work revolving around the topic of hate speech. Through the methodology of analyzing literature, we found that prolonged exposure to hate speech does mentally and physically impact annotators and researchers in this field. We also propose solutions to reduce these negative impacts such as providing mental health services, fair labor practices, psychological assessments and interventions, as well as developing AI to assist in the process of hate speech detection.

This paper introduces the Corpus of Arabic Competitive Debates (Munazarat). Despite the significance of competitive debating as an activity of fostering critical thinking and promoting dialogue, researchers within the fields of Arabic Natural Language Processing (NLP), linguistics, argumentation studies, and education have access to very limited datasets about competitive debating. At this study stage, we introduce Munazarat 1.0, which combines recordings of approximately 50 hours collected from 73 debates at QatarDebate-recognized tournaments, where all of those debates were available on YouTube. Munazarat is a novel specialized speech Arabic corpus, mostly in Modern Standard Arabic (MSA), consisting of diverse debating topics and showing rich metadata for each debate. The transcription of debates was done using Fenek, a speech-to-text Kanari AI tool, and three native Arabic speakers reviewed each transcription file to enhance the quality provided by the machine. The Munazarat 1.0 dataset can be used to train Arabic NLP tools, develop an argumentation mining machine, and analyze Arabic argumentation and rhetoric styles. Keywords: Arabic Speech Corpus, Modern Standard Arabic, Debates

On July 25, 2021, Tunisian President Kais Saied announced the suspension of parliament and dismissal of Prime Minister Hichem Mechichi, a move that sparked intense public debate. This study investigates Tunisian public opinion regarding these events by analyzing a corpus of 7,535 Facebook comments collected from the official Tunisian presidency page, specifically the post announcing the July 25 measures. A team of three annotators labeled a subset of 5,000 comments, categorizing each comment’s political stance (supportive, opposing, or neutral), sentiment (positive, negative, or neutral), emotions, presence of hate speech, aggressive tone, and racism. The inter-annotator agreement, measured by Cohen’s kappa, was 0.61, indicating substantial consensus. The analysis reveals that a majority of commenters supported President Saied’s actions, outnumbering those who opposed or took a neutral stance. Moreover, the overall sentiment expressed in the comments was predominantly positive. This study provides valuable insights into the complex landscape of public opinion in Tunisia during a crucial moment in the country’s ongoing political transformation, highlighting the role of social media as a platform for political discourse and engagement.

This study empirically investigates the role of social media in tracing the evolution of the May 2021 Israeli-Palestinian crisis, centered on the Sheikh Jarrah evictions. Analyzing a dataset of 370,747 English tweets from 120,173 users from May 9-21, 2021, the research employs a mixed-methods approach combining computational techniques and qualitative content analysis. Findings support the hypothesis that social media interactions reliably map crisis dynamics, as evidenced by hashtags like #SaveSheikhJarrah corresponding to critical shifts, though virality did not correlate with hashtag use. In contrast to prior sentiment-focused studies, the context-driven analysis reveals influencers and state actors shaping polarized narratives along geopolitical lines, with high-profile voices backing Palestinian solidarity while Israeli state accounts endorsed military operations. Evidence of a transcontinental cybercampaign emerged, albeit with limitations due to the English language scope and potential biases from data collection and keyword choices. The study contributes empirical insights into the mediatization of armed conflicts through social media’s competing narratives and information flows within the Israeli-Palestinian context. Recommendations for future multilingual, multi-platform analyses are provided to address limitations.

We present an overview of the second edition of the ArAIEval shared task, organized as part of the ArabicNLP 2024 conference co-located with ACL 2024. In this edition, ArAIEval offers two tasks: (i) detection of propagandistic textual spans with persuasion techniques identification in tweets and news articles, and (ii) distinguishing between propagandistic and non-propagandistic memes. A total of 14 teams participated in the final evaluation phase, with 6 and 9 teams participating in Tasks 1 and 2, respectively. Finally, 11 teams submitted system description papers. Across both tasks, we observed that fine-tuning transformer models such as AraBERT was at the core of the majority of the participating systems. We provide a description of the task setup, including a description of the dataset construction and the evaluation setup. We further provide a brief overview of the participating systems. All datasets and evaluation scripts are released to the research community. We hope this will enable further research on these important tasks in Arabic.

Detecting propaganda in multimodal content, such as memes, is crucial for combating disinformation on social media. This paper presents a novel approach for the ArAIEval 2024 shared Task 2 on Multimodal Propagandistic Memes Classification, involving text, image, and multimodal classification of Arabic memes. For text classification (Task 2A), we fine-tune state-of-the-art Arabic language models and use ChatGPT4-generated synthetic text for data augmentation. For image classification (Task 2B), we fine-tune ResNet18, EfficientFormerV2, and ConvNeXt-tiny architectures with DALL-E-2-generated synthetic images. For multimodal classification (Task 2C), we combine ConvNeXt-tiny and BERT architectures in a fusion layer to enhance binary classification. Our results show significant performance improvements with data augmentation for text and image classification models and with the fusion layer for multimodal classification. We highlight challenges and opportunities for future research in multimodal propaganda detection in Arabic content, emphasizing the need for robust and adaptable models to combat disinformation.

This paper focuses on detecting propagandistic spans and persuasion techniques in Arabic text from tweets and news paragraphs. Each entry in the dataset contains a text sample and corresponding labels that indicate the start and end positions of propaganda techniques within the text. Tokens falling within a labeled span were assigned ’B’ (Begin) or ’I’ (Inside) tags, ’O’, corresponding to the specific propaganda technique. Using attention masks, we created uniform lengths for each span and assigned BIO tags to each token based on the provided labels. Then, we used AraBERT-base pre-trained model for Arabic text tokenization and embeddings with a token classification layer to identify propaganda techniques. Our training process involves a two-phase fine-tuning approach. First, we train only the classification layer for a few epochs, followed by full model fine-tuning, updating all parameters. This methodology allows the model to adapt to the specific characteristics of the propaganda detection task while leveraging the knowledge captured by the pretrained AraBERT model. Our approach achieved an F1 score of 0.2774, securing the 3rd position in the leaderboard of Task 1.

We present an overview of the FIGNEWSshared task, organized as part of the Arabic-NLP 2024 conference co-located with ACL2024. The shared task addresses bias and pro-paganda annotation in multilingual news posts.We focus on the early days of the Israel War onGaza as a case study. The task aims to fostercollaboration in developing annotation guide-lines for subjective tasks by creating frame-works for analyzing diverse narratives high-lighting potential bias and propaganda. In aspirit of fostering and encouraging diversity,we address the problem from a multilingualperspective, namely within five languages: En-glish, French, Arabic, Hebrew, and Hindi. Atotal of 17 teams participated in two annota-tion subtasks: bias (16 teams) and propaganda(6 teams). The teams competed in four evalua-tion tracks: guidelines development, annotationquality, annotation quantity, and consistency.Collectively, the teams produced 129,800 datapoints. Key findings and implications for thefield are discussed.

This paper presents our team’s contribution to the FIGNEWS 2024 Shared Task, which involved annotating bias and propaganda in news coverage of the Israel-Palestine conflict. We developed comprehensive guidelines and employed a rigorous methodology to analyze 2,200 news posts from several official Facebook accounts of news websites in multiple languages. Our team, Narrative Navigators, achieved third place in both the Bias Guidelines and Bias Consistency tracks, demonstrating the effectiveness of our approach. We achieved an IAA Kappa score of 39.4 for bias annotation and 12.8 for propaganda detection. These findings and our performance underscore the need for enhanced media literacy and further research to counter the impact of biased and misleading information on public understanding of the conflict.

This paper introduces a cross-domain and multi-dialectal stance corpus for Arabic that includes four regions in the Arab World and covers the main Arabic dialect groups. Our corpus consists of 4657 sentences manually annotated with each sentence’s stance towards a specific topic. For each region, we collected sentences related to two controversial topics. We annotated each sentence by at least two annotators to indicate if its stance favors the topic, is against it, or is neutral. Our corpus is well-balanced concerning dialect and stance. Approximately half of the sentences are in Modern Standard Arabic (MSA) for each region, and the other half is in the region’s respective dialect. We conducted several machine-learning experiments for stance detection using our new corpus. Our most successful model is the Multi-Layer Perceptron (MLP), using Unigram or TF-IDF extracted features, which yielded an F1-score of 0.66 and an accuracy score of 0.66. Compared with the most similar state-of-the-art dataset, our dataset outperformed in specific stance classes, particularly “neutral” and “against”.

This paper presents the creation of the Qatari Corpus of Argumentative Writing (QCAW) as an annotated L1 Arabic and L2 English bilingual writer corpus. It comprises 200,000 tokens of argumentative writing by Qatari university students in L1 Arabic and L2 English. The corpus includes 195 essays written by 195 students, 159 females and 36 males. The students were native Arabic speakers proficient in English as a second language. The corpus is divided into Arabic and English sections, accompanied by part-of-speech annotated files. The Metadata contains information about the students (gender, major, first and second languages) and the essays (text serial numbers, word limits, genre, writing date, time spent, and location). The paper outlines the steps for collecting and analysing the corpus, including details on essay writers, topic selection, pre-analysis text modifications, proficiency level, gender, and major ratings. Statistical analyses were applied to examine the corpus. The QCAW offers a valuable bilingual data source authored by the same students in Arabic and English, with implications for further research

Social media enables widespread propagation of hate speech targeting groups based on ethnicity, religion, or other characteristics. With manual content moderation being infeasible given the volume, automatic hate speech detection is essential. This paper analyzes 70,000 Arabic tweets, from which 15,965 tweets were selected and annotated, to identify hate speech patterns and train classification models. Annotators labeled the Arabic tweets for offensive content, hate speech, emotion intensity and type, effect on readers, humor, factuality, and spam. Key findings reveal 15% of tweets contain offensive language while 6% have hate speech, mostly targeted towards groups with common ideological or political affiliations. Annotations capture diverse emotions, and sarcasm is more prevalent than humor. Additionally, 10% of tweets provide verifiable factual claims, and 7% are deemed important. For hate speech detection, deep learning models like AraBERT outperform classical machine learning approaches. By providing insights into hate speech characteristics, this work enables improved content moderation and reduced exposure to online hate. The annotated dataset advances Arabic natural language processing research and resources.

We present an overview of the ArAIEval shared task, organized as part of the first ArabicNLP 2023 conference co-located with EMNLP 2023. ArAIEval offers two tasks over Arabic text: (1) persuasion technique detection, focusing on identifying persuasion techniques in tweets and news articles, and (2) disinformation detection in binary and multiclass setups over tweets. A total of 20 teams participated in the final evaluation phase, with 14 and 16 teams participating in Task 1 and Task 2, respectively. Across both tasks, we observe that fine-tuning transformer models such as AraBERT is the core of majority of participating systems. We provide a description of the task setup, including description of datasets construction and the evaluation setup. We also provide a brief overview of the participating systems. All datasets and evaluation scripts from the shared task are released to the research community. We hope this will enable further research on such important tasks within the Arabic NLP community.

Propaganda is defined as an expression of opinion or action by individuals or groups deliberately designed to influence opinions or actions of other individuals or groups with reference to predetermined ends and this is achieved by means of well-defined rhetorical and psychological devices. Currently, propaganda (or persuasion) techniques have been commonly used on social media to manipulate or mislead social media users. Automatic detection of propaganda techniques from textual, visual, or multimodal content has been studied recently, however, major of such efforts are focused on English language content. In this paper, we propose a shared task on detecting propaganda techniques for Arabic textual content. We have done a pilot annotation of 200 Arabic tweets, which we plan to extend to 2,000 tweets, covering diverse topics. We hope that the shared task will help in building a community for Arabic propaganda detection. The dataset will be made publicly available, which can help in future studies.

The task of machine reading comprehension (MRC) is a useful benchmark to evaluate the natural language understanding of machines. It has gained popularity in the natural language processing (NLP) field mainly due to the large number of datasets released for many languages. However, the research in MRC has been understudied in several domains, including religious texts. The goal of the Qur’an QA 2022 shared task is to fill this gap by producing state-of-the-art question answering and reading comprehension research on Qur’an. This paper describes the DTW entry to the Quran QA 2022 shared task. Our methodology uses transfer learning to take advantage of available Arabic MRC data. We further improve the results using various ensemble learning strategies. Our approach provided a partial Reciprocal Rank (pRR) score of 0.49 on the test set, proving its strong performance on the task.

This paper describes our participation in the shared task Fine-Grained Hate Speech Detection on Arabic Twitter at the 5th Workshop on Open-Source Arabic Corpora and Processing Tools (OSACT). The shared task is divided into three detection subtasks: (i) Detect whether a tweet is offensive or not; (ii) Detect whether a tweet contains hate speech or not; and (iii) Detect the fine-grained type of hate speech (race, religion, ideology, disability, social class, and gender). It is an effort toward the goal of mitigating the spread of offensive language and hate speech in Arabic-written content on social media platforms. To solve the three subtasks, we employed six different transformer versions: AraBert, AraElectra, Albert-Arabic, AraGPT2, mBert, and XLM-Roberta. We experimented with models based on encoder and decoder blocks and models exclusively trained on Arabic and also on several languages. Likewise, we applied two ensemble methods: Majority vote and Highest sum. Our approach outperformed the official baseline in all the subtasks, not only considering F1-macro results but also accuracy, recall, and precision. The results suggest that the Highest sum is an excellent approach to encompassing transformer output to create an ensemble since this method offered at least top-two F1-macro values across all the experiments performed on development and test data.

We present the results and the main findings of the NLP4IF-2021 shared tasks. Task 1 focused on fighting the COVID-19 infodemic in social media, and it was offered in Arabic, Bulgarian, and English. Given a tweet, it asked to predict whether that tweet contains a verifiable claim, and if so, whether it is likely to be false, is of general interest, is likely to be harmful, and is worthy of manual fact-checking; also, whether it is harmful to society, and whether it requires the attention of policy makers. Task 2 focused on censorship detection, and was offered in Chinese. A total of ten teams submitted systems for task 1, and one team participated in task 2; nine teams also submitted a system description paper. Here, we present the tasks, analyze the results, and discuss the system submissions and the methods they used. Most submissions achieved sizable improvements over several baselines, and the best systems used pre-trained Transformers and ensembles. The data, the scorers and the leaderboards for the tasks are available at http://gitlab.com/NLP4IF/nlp4if-2021.

This paper provides an overview of the WANLP 2021 shared task on sarcasm and sentiment detection in Arabic. The shared task has two subtasks: sarcasm detection (subtask 1) and sentiment analysis (subtask 2). This shared task aims to promote and bring attention to Arabic sarcasm detection, which is crucial to improve the performance in other tasks such as sentiment analysis. The dataset used in this shared task, namely ArSarcasm-v2, consists of 15,548 tweets labelled for sarcasm, sentiment and dialect. We received 27 and 22 submissions for subtasks 1 and 2 respectively. Most of the approaches relied on using and fine-tuning pre-trained language models such as AraBERT and MARBERT. The top achieved results for the sarcasm detection and sentiment analysis tasks were 0.6225 F1-score and 0.748 F1-PN respectively.

With the emergence of the COVID-19 pandemic, the political and the medical aspects of disinformation merged as the problem got elevated to a whole new level to become the first global infodemic. Fighting this infodemic has been declared one of the most important focus areas of the World Health Organization, with dangers ranging from promoting fake cures, rumors, and conspiracy theories to spreading xenophobia and panic. Addressing the issue requires solving a number of challenging problems such as identifying messages containing claims, determining their check-worthiness and factuality, and their potential to do harm as well as the nature of that harm, to mention just a few. To address this gap, we release a large dataset of 16K manually annotated tweets for fine-grained disinformation analysis that (i) focuses on COVID-19, (ii) combines the perspectives and the interests of journalists, fact-checkers, social media platforms, policy makers, and society, and (iii) covers Arabic, Bulgarian, Dutch, and English. Finally, we show strong evaluation results using pretrained Transformers, thus confirming the practical utility of the dataset in monolingual vs. multilingual, and single task vs. multitask settings.

The current Arabic natural language processing resources are mainly build to address the Modern Standard Arabic (MSA), while we witnessed some scattered efforts to build resources for various Arabic dialects such as the Levantine and the Egyptian dialects. We observed a lack of resources for Gulf Arabic and especially the Qatari variety. In this paper, we present the first Qatari idioms and expression corpus of 1000 entries. The corpus was created from on-line and printed sources in addition to transcribed recorded interviews. The corpus covers various Qatari traditional expressions and idioms. To this end, audio recordings were collected from interviews and an online survey questionnaire was conducted to validate our data. This corpus aims to help advance the dialectal Arabic Speech and Natural Language Processing tools and applications for the Qatari dialect.

Identifying irony in user-generated social media content has a wide range of applications; however to date Arabic content has received limited attention. To bridge this gap, this study builds a new open domain Arabic corpus annotated for irony detection. We query Twitter using irony-related hashtags to collect ironic messages, which are then manually annotated by two linguists according to our working definition of irony. Challenges which we have encountered during the annotation process reflect the inherent limitations of Twitter messages interpretation, as well as the complexity of Arabic and its dialects. Once published, our corpus will be a valuable free resource for developing open domain systems for automatic irony recognition in Arabic language and its dialects in social media text.

We present ARAP-Tweet 2.0, a corpus of 5 million dialectal Arabic tweets and 50 million words of about 3000 Twitter users from 17 Arab countries. Compared to the first version, the new corpus has significant improvements in terms of the data volume and the annotation quality. It is fully balanced with respect to dialect, gender, and three age groups: under 25 years, between 25 and 34, and 35 years and above. This paper describes the process of creating the corpus starting from gathering the dialectal phrases to find the users, to annotating their accounts and retrieving their tweets. We also report on the evaluation of the annotation quality using the inter-annotator agreement measures which were applied to the whole corpus and not just a subset. The obtained results were substantial with average Cohen’s Kappa values of 0.99, 0.92, and 0.88 for the annotation of gender, dialect, and age respectively. We also discuss some challenges encountered when developing this corpus.s.

Arabic writing is typically underspecified for short vowels and other markups, referred to as diacritics. In addition to the lexical ambiguity exhibited in most languages, the lack of diacritics in written Arabic adds another layer of ambiguity which is an artifact of the orthography. In this paper, we present the details of three annotation experimental conditions designed to study the impact of automatic ambiguity detection, on annotation speed and quality in a large scale annotation project.

We present our guidelines and annotation procedure to create a human corrected machine translated post-edited corpus for the Modern Standard Arabic. Our overarching goal is to use the annotated corpus to develop automatic machine translation post-editing systems for Arabic that can be used to help accelerate the human revision process of translated texts. The creation of any manually annotated corpus usually presents many challenges. In order to address these challenges, we created comprehensive and simplified annotation guidelines which were used by a team of five annotators and one lead annotator. In order to ensure a high annotation agreement between the annotators, multiple training sessions were held and regular inter-annotator agreement measures were performed to check the annotation quality. The created corpus of manual post-edited translations of English to Arabic articles is the largest to date for this language pair.

This paper presents the annotation guidelines developed as part of an effort to create a large scale manually diacritized corpus for various Arabic text genres. The target size of the annotated corpus is 2 million words. We summarize the guidelines and describe issues encountered during the training of the annotators. We also discuss the challenges posed by the complexity of the Arabic language and how they are addressed. Finally, we present the diacritization annotation procedure and detail the quality of the resulting annotations.

The goal of the cognitive machine translation (MT) evaluation approach is to build classifiers which assign post-editing effort scores to new texts. The approach helps estimate fair compensation for post-editors in the translation industry by evaluating the cognitive difficulty of post-editing MT output. The approach counts the number of errors classified in different categories on the basis of how much cognitive effort they require in order to be corrected. In this paper, we present the results of applying an existing cognitive evaluation approach to Modern Standard Arabic (MSA). We provide a comparison of the number of errors and categories of errors in three MSA texts of different MT quality (without any language-specific adaptation), as well as a comparison between MSA texts and texts from three Indo-European languages (Russian, Spanish, and Bulgarian), taken from a previous experiment. The results show how the error distributions change passing from the MSA texts of worse MT quality to MSA texts of better MT quality, as well as a similarity in distinguishing the texts of better MT quality for all four languages.

Crowdsourcing has been used recently as an alternative to traditional costly annotation by many natural language processing groups. In this paper, we explore the use of Amazon Mechanical Turk (AMT) in order to assess the feasibility of using AMT workers (also known as Turkers) to perform linguistic annotation of Arabic. We used a gold standard data set taken from the Quran corpus project annotated with part-of-speech and morphological information. An Arabic language qualification test was used to filter out potential non-qualified participants. Two experiments were performed, a part-of-speech tagging task in where the annotators were asked to choose a correct word-category from a multiple choice list and case ending identification task. The results obtained so far showed that annotating Arabic grammatical case is harder than POS tagging, and crowdsourcing for Arabic linguistic annotation requiring expert annotators could be not as effective as other crowdsourcing experiments requiring less expertise and qualifications.

We present annotation guidelines and a web-based annotation framework developed as part of an effort to create a manually annotated Arabic corpus of errors and corrections for various text types. Such a corpus will be invaluable for developing Arabic error correction tools, both for training models and as a gold standard for evaluating error correction algorithms. We summarize the guidelines we created. We also describe issues encountered during the training of the annotators, as well as problems that are specific to the Arabic language that arose during the annotation process. Finally, we present the annotation tool that was developed as part of this project, the annotation pipeline, and the quality of the resulting annotations.

Dans cette démonstration, nous présentons l’implémentation d’un outil de repérage d’entités nommées à base de règle pour la langue arabe dans le système de veille médiatique EMM (Europe Media Monitor).

The Arabic Treebank (ATB) Project at the Linguistic Data Consortium (LDC) has embarked on a large corpus of Broadcast News (BN) transcriptions, and this has led to a number of new challenges for the data processing and annotation procedures that were originally developed for Arabic newswire text (ATB1, ATB2 and ATB3). The corpus requirements currently posed by the DARPA GALE Program, including English translation of Arabic BN transcripts, word-level alignment of Arabic and English data, and creation of a corresponding English Treebank, place significant new constraints on ATB corpus creation, and require careful coordination among a wide assortment of concurrent activities and participants. Nonetheless, in spite of the new challenges posed by BN data, the ATB’s newly improved pipeline and revised annotation guidelines for newswire have proven to be robust enough that very few changes were necessary to account for the new genre of data. This paper presents the points where some adaptation has been necessary, and the overall pipeline as used in the production of BN ATB data.

We present a fully functional Arabic information extraction (IE) system that is used to analyze large volumes of news texts every day to extract the named entity (NE) types person, organization, location, date and number, as well as quotations (direct reported speech) by and about people. The Named Entity Recognition (NER) system was not developed for Arabic, but - instead - a highly multilingual, almost language-independent NER system was adapted to also cover Arabic. The Semitic language Arabic substantially differs from the Indo-European and Finno-Ugric languages currently covered. This paper thus describes what Arabic language-specific resources had to be developed and what changes needed to be made to the otherwise language-independent rule set in order to be applicable to the Arabic language. The achieved evaluation results are generally satisfactory, but could be improved for certain entity types. The results of the IE tools can be seen on the Arabic pages of the freely accessible Europe Media Monitor (EMM) application NewsExplorer, which can be found at http://press.jrc.it/overview.html.

In this paper, we present the details of creating a pilot Arabic proposition bank (Propbank). Propbanks exist for both English and Chinese. However the morphological and syntactic expression of linguistic phenomena in Arabic yields a very different type of process in creating an Arabic propbank. Hence, we highlight those characteristics of Arabic that make creating a propbank for the language a different challenge compared to the creation of an English Propbank.We believe that many of the lessons learned in dealing with Arabic could generalise to other languages that exhibit equally rich morphology and relatively free word order.

This paper describes the ARCADE II project, concerned with the evaluation of parallel text alignment systems. The ARCADE II project aims at exploring the techniques of multilingual text alignment through a fine evaluation of the existing techniques and the development of new alignment methods. The evaluation campaign consists of two tracks devoted to the evaluation of alignment at sentence and word level respectively. It differs from ARCADE I in the multilingual aspect and the investigation of lexical alignment.

We are presenting a method to recognise geographical references in free text. Our tool must work on various languages with a minimum of language-dependent resources, except a gazetteer. The main difficulty is to disambiguate these place names by distinguishing places from persons and by selecting the most likely place out of a list of homographic place names world-wide. The system uses a number of language-independent clues and heuristics to disambiguate place name homographs. The final aim is to index texts with the countries and cities they mention and to automatically visualise this information on geographical maps using various tools.