The recent development and success of Large Language Models (LLMs) necessitate an evaluation of their performance across diverse NLP tasks in different languages. Although several frameworks have been developed and made publicly available, their customization capabilities for specific tasks and datasets are often complex for different users. In this study, we introduce the LLMeBench framework, which can be seamlessly customized to evaluate LLMs for any NLP task, regardless of language. The framework features generic dataset loaders, several model providers, and pre-implements most standard evaluation metrics. It supports in-context learning with zero- and few-shot settings. A specific dataset and task can be evaluated for a given LLM in less than 20 lines of code while allowing full flexibility to extend the framework for custom datasets, models, or tasks. The framework has been tested on 31 unique NLP tasks using 53 publicly available datasets within 90 experimental setups, involving approximately 296K data points. We open-sourced LLMeBench for the community (https://github.com/qcri/LLMeBench/) and a video demonstrating the framework is available online (https://youtu.be/9cC2m_abk3A).

The recent breakthroughs in Artificial Intelligence (AI) can be attributed to the remarkable performance of Large Language Models (LLMs) across a spectrum of research areas (e.g., machine translation, question-answering, automatic speech recognition, text-to-speech generation) and application domains (e.g., business, law, healthcare, education, and psychology). The success of these LLMs largely de- pends on specific training techniques, most notably instruction tuning, RLHF, and subsequent prompting to achieve the desired output. As the development of such LLMs continues to increase in both closed and open settings, evaluation has become crucial for understanding their generalization capabilities across different tasks, modalities, languages, and dialects. This evaluation process is tightly coupled with prompting, which plays a key role in obtain- ing better outputs. There has been attempts to evaluate such models focusing on diverse tasks, languages, and dialects, which suggests that the capabilities of LLMs are still limited to medium-to-low-resource languages due to the lack of representative datasets. The tutorial offers an overview of this emerging research area. We explore the capabilities of LLMs in terms of their performance, zero- and few-shot settings, fine-tuning, instructions tuning, and close vs. open models with a special emphasis on low-resource settings. In addition to LLMs for standard NLP tasks, we will focus on speech and multimodality.

The emergence of the COVID-19 pandemic and the first global infodemic have changed our lives in many different ways. We relied on social media to get the latest information about COVID-19 pandemic and at the same time to disseminate information. The content in social media consisted not only health related advice, plans, and informative news from policymakers, but also contains conspiracies and rumors. It became important to identify such information as soon as they are posted to make an actionable decision (e.g., debunking rumors, or taking certain measures for traveling). To address this challenge, we develop and publicly release the first largest manually annotated Arabic tweet dataset, ArCovidVac, for COVID-19 vaccination campaign, covering many countries in the Arab region. The dataset is enriched with different layers of annotation, including, (i) Informativeness more vs. less importance of the tweets); (ii) fine-grained tweet content types (e.g., advice, rumors, restriction, authenticate news/information); and (iii) stance towards vaccination (pro-vaccination, neutral, anti-vaccination). Further, we performed in-depth analysis of the data, exploring the popularity of different vaccines, trending hashtags, topics, and presence of offensiveness in the tweets. We studied the data for individual types of tweets and temporal changes in stance towards vaccine. We benchmarked the ArCovidVac dataset using transformer architectures for informativeness, content types, and stance detection.

Gender analysis of Twitter can reveal important socio-cultural differences between male and female users. There has been a significant effort to analyze and automatically infer gender in the past for most widely spoken languages’ content, however, to our knowledge very limited work has been done for Arabic. In this paper, we perform an extensive analysis of differences between male and female users on the Arabic Twitter-sphere. We study differences in user engagement, topics of interest, and the gender gap in professions. Along with gender analysis, we also propose a method to infer gender by utilizing usernames, profile pictures, tweets, and networks of friends. In order to do so, we manually annotated gender and locations for ~166K Twitter accounts associated with ~92K user location, which we plan to make publicly available. Our proposed gender inference method achieve an F1 score of 82.1% (47.3% higher than majority baseline). We also developed a demo and made it publicly available.

We introduce the largest transcribed Arabic speech corpus, QASR, collected from the broadcast domain. This multi-dialect speech dataset contains 2,000 hours of speech sampled at 16kHz crawled from Aljazeera news channel. The dataset is released with lightly supervised transcriptions, aligned with the audio segments. Unlike previous datasets, QASR contains linguistically motivated segmentation, punctuation, speaker information among others. QASR is suitable for training and evaluating speech recognition systems, acoustics- and/or linguistics- based Arabic dialect identification, punctuation restoration, speaker identification, speaker linking, and potentially other NLP modules for spoken data. In addition to QASR transcription, we release a dataset of 130M words to aid in designing and training a better language model. We show that end-to-end automatic speech recognition trained on QASR reports a competitive word error rate compared to the previous MGB-2 corpus. We report baseline results for downstream natural language processing tasks such as named entity recognition using speech transcript. We also report the first baseline for Arabic punctuation restoration. We make the corpus available for the research community.

In this paper, we describe our efforts at OSACT Shared Task on Offensive Language Detection. The shared task consists of two subtasks: offensive language detection (Subtask A) and hate speech detection (Subtask B). For offensive language detection, a system combination of Support Vector Machines (SVMs) and Deep Neural Networks (DNNs) achieved the best results on development set, which ranked 1st in the official results for Subtask A with F1-score of 90.51% on the test set. For hate speech detection, DNNs were less effective and a system combination of multiple SVMs with different parameters achieved the best results on development set, which ranked 4th in official results for Subtask B with F1-macro score of 80.63% on the test set.

Access to social media often enables users to engage in conversation with limited accountability. This allows a user to share their opinions and ideology, especially regarding public content, occasionally adopting offensive language. This may encourage hate crimes or cause mental harm to targeted individuals or groups. Hence, it is important to detect offensive comments in social media platforms. Typically, most studies focus on offensive commenting in one platform only, even though the problem of offensive language is observed across multiple platforms. Therefore, in this paper, we introduce and make publicly available a new dialectal Arabic news comment dataset, collected from multiple social media platforms, including Twitter, Facebook, and YouTube. We follow two-step crowd-annotator selection criteria for low-representative language annotation task in a crowdsourcing platform. Furthermore, we analyze the distinctive lexical content along with the use of emojis in offensive comments. We train and evaluate the classifiers using the annotated multi-platform dataset along with other publicly available data. Our results highlight the importance of multiple platform dataset for (a) cross-platform, (b) cross-domain, and (c) cross-dialect generalization of classifier performance.

Automatic categorization of short texts, such as news headlines and social media posts, has many applications ranging from content analysis to recommendation systems. In this paper, we use such text categorization i.e., labeling the social media posts to categories like ‘sports’, ‘politics’, ‘human-rights’ among others, to showcase the efficacy of models across different sources and varieties of Arabic. In doing so, we show that diversifying the training data, whether by using diverse training data for the specific task (an increase of 21% macro F1) or using diverse data to pre-train a BERT model (26% macro F1), leads to overall improvements in classification effectiveness. In our work, we also introduce two new Arabic text categorization datasets, where the first is composed of social media posts from a popular Arabic news channel that cover Twitter, Facebook, and YouTube, and the second is composed of tweets from popular Arabic accounts. The posts in the former are nearly exclusively authored in modern standard Arabic (MSA), while the tweets in the latter contain both MSA and dialectal Arabic.

We propose a novel approach to the study of how artificial neural network perceive the distinction between grammatical and ungrammatical sentences, a crucial task in the growing field of synthetic linguistics. The method is based on performance measures of language models trained on corpora and fine-tuned with either grammatical or ungrammatical sentences, then applied to (different types of) grammatical or ungrammatical sentences. The results show that both in the difficult and highly symmetrical task of detecting subject islands and in the more open CoLA dataset, grammatical sentences give rise to better scores than ungrammatical ones, possibly because they can be better integrated within the body of linguistic structural knowledge that the language model has accumulated.

The paper explores the ability of LSTM networks trained on a language modeling task to detect linguistic structures which are ungrammatical due to extraction violations (extra arguments and subject-relative clause island violations), and considers its implications for the debate on language innatism. The results show that the current RNN model can correctly classify (un)grammatical sentences, in certain conditions, but it is sensitive to linguistic processing factors and probably ultimately unable to induce a more abstract notion of grammaticality, at least in the domain we tested.

Silence is an integral part of the most frequent turn-taking phenomena in spoken conversations. Silence is sized and placed within the conversation flow and it is coordinated by the speakers along with the other speech acts. The objective of this analytical study is twofold: to explore the functions of silence with duration of one second and above, towards information flow in a dyadic conversation utilizing the sequences of dialog acts present in the turns surrounding the silence itself; and to design a feature space useful for clustering the silences using a hierarchical concept formation algorithm. The resulting clusters are manually grouped into functional categories based on their similarities. It is observed that the silence plays an important role in response preparation, also can indicate speakers’ hesitation or indecisiveness. It is also observed that sometimes long silences can be used deliberately to get a forced response from another speaker thus making silence a multi-functional and an important catalyst towards information flow.

Spoken conversation corpora often adapt existing Dialogue Act (DA) annotation specifications, such as DAMSL, DIT++, etc., to task specific needs, yielding incompatible annotations; thus, limiting corpora re-usability. Recently accepted ISO standard for DA annotation – Dialogue Act Markup Language (DiAML) – is designed as domain and application independent. Moreover, the clear separation of dialogue dimensions and communicative functions, coupled with the hierarchical organization of the latter, allows for classification at different levels of granularity. However, re-annotating existing corpora with the new scheme might require significant effort. In this paper we test the utility of the ISO standard through comparative evaluation of the corpus-specific legacy and the semi-automatically transferred DiAML DA annotations on supervised dialogue act classification task. To test the domain independence of the resulting annotations, we perform cross-domain and data aggregation evaluation. Compared to the legacy annotation scheme, on the Italian LUNA Human-Human corpus, the DiAML annotation scheme exhibits better cross-domain and data aggregation classification performance, while maintaining comparable in-domain performance.

In this paper, we aim to investigate the coordination of interlocutors behavior in different emotional segments. Conversational coordination between the interlocutors is the tendency of speakers to predict and adjust each other accordingly on an ongoing conversation. In order to find such a coordination, we investigated 1) lexical similarities between the speakers in each emotional segments, 2) correlation between the interlocutors using psycholinguistic features, such as linguistic styles, psychological process, personal concerns among others, and 3) relation of interlocutors turn-taking behaviors such as competitiveness. To study the degree of coordination in different emotional segments, we conducted our experiments using real dyadic conversations collected from call centers in which agent’s emotional state include empathy and customer’s emotional states include anger and frustration. Our findings suggest that the most coordination occurs between the interlocutors inside anger segments, where as, a little coordination was observed when the agent was empathic, even though an increase in the amount of non-competitive overlaps was observed. We found no significant difference between anger and frustration segment in terms of turn-taking behaviors. However, the length of pause significantly decreases in the preceding segment of anger where as it increases in the preceding segment of frustration.