Intention detection is a crucial aspect of natural language understanding (NLU), focusing on identifying the primary objective underlying user input. In this work, we present a transformer-based method that excels in determining the intent of Arabic text within the banking domain. We explored several machine learning (ML), deep learning (DL), and transformer-based models on an Arabic banking dataset for intent detection. Our findings underscore the challenges that traditional ML and DL models face in understanding the nuances of various Arabic dialects, leading to subpar performance in intent detection. However, the transformer-based methods, designed to tackle such complexities, significantly outperformed the other models in classifying intent across different Arabic dialects. Notably, the AraBERTv2 model achieved the highest micro F1 score of 82.08% in ArBanking77 dataset, a testament to its effectiveness in this context. This achievement, which contributed to our work being ranked 5$^{th}$ in the shared task, AraFinNLP2024, highlights the importance of developing models that can effectively handle the intricacies of Arabic language processing and intent detection.

SemanticCuetSync at AraFinNLP2024: Classification of Cross-Dialect Intent in the Banking Domain using Transformers

Detecting propagandistic spans and identifying persuasion techniques are crucial for promoting informed decision-making, safeguarding democratic processes, and fostering a media environment characterized by integrity and transparency. Various machine learning (Logistic Regression, Random Forest, and Multinomial Naive Bayes), deep learning (CNN, CNN+LSTM, CNN+BiLSTM), and transformer-based (AraBERTv2, AraBERT-NER, CamelBERT, BERT-Base-Arabic) models were exploited to perform the task. The evaluation results indicate that CamelBERT achieved the highest micro-F1 score (24.09%), outperforming CNN+LSTM and AraBERTv2. The study found that most models struggle to detect propagandistic spans when multiple spans are present within the same article. Overall, the model's performance secured a $6^{th}$ place ranking in the ArAIEval Shared Task-1.

SemanticCuetSync at ArAIEval Shared Task: Detecting Propagandistic Spans with Persuasion Techniques Identification using Pre-trained Transformers

Hate and offensive language in online platforms pose significant challenges, necessitating automatic detection methods. Particularly in the case of codemixed text, which is very common in social media, the complexity of this problem increases due to the cultural nuances of different languages. A shared task on the detection of hate and offensive language for Telugu was organized by DravidianLangTech. To complete this task, our study investigates the effectiveness of transliteration-augmented datasets for Telugu codemixed text. We compare the performance of various machine learning, deep learning, and transformer-based models on both original and augmented datasets. Our findings demonstrate the superiority of transformer models, particularly Telugu-BERT, which achieved an $f_1$ score of 0.77 on the augmented dataset, ranking the $1^{st}$ position in the leaderboard. The study highlights the potential of transliteration-augmented datasets in improving model performance and suggests further exploration of diverse transliteration options to address real-world scenarios.

Sandalphon@DravidianLangTech-EACL2024: Hate and Offensive Language Detection in Telugu Codemixed Text using Transliteration-Augmentation

Due to technological advancements, various methods have emerged for disseminating news to the masses. The pervasive reach of news, however, has given rise to a significant concern - the proliferation of fake news. In response to this challenge, a shared task in DravidianLangTech@EACL 2024 was initiated to detect fake news and classify its types in the Malayalam language. The shared task consisted of two sub-tasks. Task-1 focused on a binary classification problem, determining whether a piece of news is fake or not. Whereas task-2 delved into a multi-class classification problem, categorizing news into five distinct levels. Our approach involved the exploration of various machine learning, deep learning, and transformer models, with an emphasis on parameter tuning to enhance overall model performance. As a result, we introduce a MuRIL model that leverages parameter tuning, achieving notable success with an F1-score of 0.86 in task-1 and 0.5191 in task-2. This successful implementation led to our system securing the 3rd position in task-1 and the 1st position in task-2. The source code will be found in the GitHub repository provided in this link: https://github.com/Salman1804102/DravidianLangTech-EACL-2024-FakeNews.

CUET_Binary_Hackers@DravidianLangTech@EACL 2024: Malayalam Fake News Detection by Fine-Tuned MuRIL BERT

Fake news on social media is increasingly used to target and manipulate public opinion, influencing the decisions of a large number of people beyond mere misinformation. Researchers are increasingly interested in verifying news legitimacy due to the difficulty and cost of manual verification, especially with the increasing use of social media and internet access. Unfortunately, research on low-resource languages like Malayalam remains negligible due to a lack of resources and language processing tools. To address the issue the paper describes the submission of the team Punny Punctuators to the shared task on Fake News Detection in Dravidian Languages-DravidianLangTech@EACL 2024. The shared task consists of two sub-tasks: (i) the first subtask classifies a given social media text as original or fake, and (ii) the second subtask classifies fake news into different categories. We used many transformer-based, deep-learning, and machine-learning models. The Malayalam-BERT model performs better than the other methods in subtask-1 and subtask-2, placing it 2nd with a macro F1-score of 0.87 and 11th with a macro F1-score of 0.17, respectively.

Punny_Punctuators@DravidianLangTech-EACL2024: Detection and Classification of Fake News for Malayalam Social-Media Text

In this modern era, many people have been using Facebook and Twitter, leading to a rise in information sharing and communication. However, a considerable amount of information on these platforms is misleading or intentionally crafted to deceive users, which is often termed "Fake News". This shared task on Fake News Detection in Dravidian Languages (Task-1) organized by DravidianLangTech@EACL 2024 allows for addressing the challenge of distinguishing between original and fake news content in the Malayalam language. Our approach involves creating an intelligent model to categorize text as either fake or original. We experimented with various models, including Logistic Regression, Decision Tree, Random Forest, SVM, and SGD for machine learning; CNN, BiLSTM, and BiLSTM with attention to deep learning; and XLM-R and m-BERT for transformer-based approaches. Notably, our most successful model, m-BERT, achieved a top-ranking position with a macro F1 score of 0.85, ranked 4th among the participants. This research contributes to combating misinformation on social media news, offering an effective solution to classify content accurately.

CUET_NLP_GoodFellows@DravidianLangTech 2024: A transformer-based approach for detecting Fake News in Dravidian Languages

With the continuous evolution of technology and widespread internet access, various social media platforms have gained immense popularity, attracting a vast number of active users globally. However, this surge in online activity has also led to a concerning trend by driving many individuals to resort to posting hateful and offensive comments or posts, publicly targeting groups or individuals. In response to these challenges, we participated in this shared task. Our approach involved proposing a fine-tuning-based pre-trained transformer model to effectively discern whether a given text contains offensive content that propagates hatred. We conducted comprehensive experiments, exploring various machine learning (LR, SVM, and Ensemble), deep learning (CNN, BiLSTM, CNN+BiLSTM), and transformer models (indic-SBERT, mBERT, MuRIL, Distil-mBERT, XLMR), adhering to a meticulous fine-tuning methodology. Among the models evaluated, our fine-tuned L3Cube-Indic-Sentence-Similarity-BERT or Indic-SBERT model demonstrated superior performance, achieving a macro-average F1-score of 0.7013. This notable result positioned us at the 6th place in the task. The implementation details of the task will be found in the GitHub repository.

CUET_Binary_Hackers@DravidianLangTech@EACL 2024: Hate and Offensive Language Detection in Telugu Code-Mixed Text Using Sentence Similarity BERT

Sentiment analysis delves into people's opinions, intuitions, and emotions regarding any entity. Within natural language processing (NLP), it serves as a technique to extract subjective knowledge, determining whether an opinion or comment leans positive, negative, neutral, or a mix thereof towards an entity. In recent years, it has garnered substantial attention from NLP researchers due to the vast availability of comments and opinions on the internet. Despite extensive studies in this domain, sentiment analysis in low-resource languages such as Tamil and Tulu faces challenges in handling code-mixed and transliterated content. To address these challenges, this work focuses on sentiment analysis of code-mixed and transliterated Tamil and Tulu social media comments. It explores four machine learning approaches (LR, SVM, XGBoost, Ensemble), two deep learning methods (BiLSTM with fastText and Word2Vec), and four transformer-based models (m-BERT, MuRIL, L3Cube-IndicSBERT, and DistilmBERT) for both languages. For Tamil, L3Cube-IndicSBERT and ensemble approaches outperform others, while m-BERT demonstrates superior performance among the models for Tulu. The presented models achieved the third and first ranks by attaining macro F1-scores of 0.227 and 0.584 in Tamil and Tulu, respectively.

CUET_Binary_Hackers@DravidianLangTech-EACL 2024: Sentiment Analysis using Transformer-Based Models in Code-Mixed and Transliterated Tamil and Tulu

Abusive language is a major problem on social media, which highlights the need for efficient detection methods to be developed. This research addresses the formidable challenge of detecting abusive language in Tamil through a comprehensive multimodal approach, incorporating textual, acoustic, and visual inputs. In the domain of video data, our study utilizes ConvLSTM, 3DCNN, and a hybrid 3DCNN with BiLSTM. For audio data, models such as LSTM, BiLSTM, RNN, LR, GNB, and RF are explored. Regarding textual content, we investigate MNB, LR, RF, RNN, LSTM, and BiLSTM. To further enhance overall performance, we introduce a weighted late fusion model amalgamating predictions from all modalities. The fusion model is then applied to make predictions on our test dataset, yielding macro f1 score of 71.43%. Within the shared task "Multimodal Abusive Language Detection and Sentiment Analysis in Dravidian Languages: DravidianLangTech@EACL 2024," this methodology has achieved the highest ranking for detecting abusive language detection in the Tamil.

Binary_Beasts@DravidianLangTech@EACL 2024: Multimodal Abusive Language Detection in Tamil: An Integrated Approach using ML and DL Techniques

The importance of being able to identify between fake and original news in the arena of social media demands surveillance procedures. This paper simply establishes the shared task on 'Fake News Detection in Dravidian Languages - DravidianLangTech@EACL 2024.' Focused on the Malayalam language, the task entails identifying social media posts as either fake or original news. The participating teams presented varied strategies, ranging from conventional machine-learning techniques to cutting-edge transformer-based models. Notably, our findings underscore the usefulness of the Malayalam-BERT model, demonstrating a noteworthy accuracy of 88% in detecting between fake and original news among Malayalam social media material.

CUET_DUO@DravidianLangTech 2024: Fake News Classification Using Malayalam-BERT

Social media reviews are utilized in sentiment analysis (SA), a computing area that analyses people's thoughts, and attitudes which has increased the need for efficient tools to analyze and determine feelings in such content. To tackle this challenge, our study introduces a system developed for the shared task of analyzing sentiment in Tamil and Tulu code-mixed languages. We utilized a combination of ML models (DT, RF, MNB, SVM), DL models (CNN, BiLSTM, CNN+BiLSTM with FastText), and transformer-based models (Indic-BERT, XLM-RoBERTa, m-BERT) for both Tamil and Tulu code-mixed language. Experimental outcomes reveal that the transformer-based models XLM-RoBERTa and m-BERT surpassed others in performance for Tamil and Tulu, respectively. The proposed XLM-RoBERTa and m-BERT models attained macro F1-scores of 0.258 and 0.468 for the Tamil and Tulu datasets, securing the 2nd and 5th positions, respectively, in the shared task competition.

CUETSentimentSillies@DravidianLangTech-EACL2024: Transformer-based approach for Sentiment Analysis in Tamil and Tulu Code-Mixed Text

False news misleads people and may lead to real-world miscommunication and injury. Identifying fake news teaches society. Removing misinformation encourages critical thinking, democracy, and the prevention of hatred, fear, and misunderstanding. Identifying and removing fake news and developing a detection system is essential for reliable, accurate, and clear information. Therefore a shared task is organized to detect fake news in Malayalam. This paper presents a system developed for the shared task of detecting and classifying fake news in Malayalam. The approach involves a combination of machine learning models(DT, RF, MNB, SVM), deep learning models (CNN, BiLSTM, CNN+BiLSTM with FastText), and transformer-based models (Indic-BERT, indic-BERT, m-BERT) for both subtasks. The experimental results demonstrate that transformer-based models, specifically m-BERT and Malayalam-BERT, outperformed others. They achieved superior performance with noteworthy macro F1-scores of 0.84 and 0.496 for subtask 1 and subtask 2, respectively, securing the 5th and 2nd positions in the shared task competition.

CUETSentimentSillies@DravidianLangTech-EACL2024: Transformer-based approach for Detecting and Categorizing Fake News in Malayalam Language

The widespread use of online communication has caused a significant increase in the spread of hate speech on social media. However, there are also hate crimes based on caste or immigration status. Despite India's efforts to bring equality among its citizens, numerous crimes occur just based on caste. Immigration-based hostility happens both in India and in developed countries. A shared task was arranged to address this issue in a low-resource language such as Tamil. This paper aims to improve the detection of hate speech and hostility based on caste and immigration status on social media. To achieve this, we investigated several Machine Learning (ML), Deep Learning (DL), and transformer-based models, including M-BERT, XLM-R, and Tamil BERT. We obtained a macro -score of 0.80 using the M-BERT model, which enabled us to rank on the shared task.

CUET_NLP_Manning_LT-EDI@EACL2024: Transformer Based Approach on Caste or Immigration Hate Speech Detection

The pervasive impact of stress on individuals necessitates proactive identification and intervention measures, especially in social media interaction. This research paper addresses the imperative need for proactive identification and intervention concerning the widespread influence of stress on individuals. This study focuses on the shared task, "Stress Identification in Dravidian Languages," specifically emphasizing Tamil and Telugu code-mixed languages. The primary objective of the task is to classify social media messages into two categories: stressed and non stressed. We employed various methodologies, from traditional machine-learning techniques to state-of-the-art transformer-based models. Notably, the Tamil-BERT and Telugu-BERT models exhibited exceptional performance, achieving a noteworthy macro F1-score of \textbf{0.71} and \textbf{0.72}, respectively, and securing the position in Tamil code-mixed language and the position in the Telugu code-mixed language. These findings underscore the effectiveness of these models in recognizing stress signals within social media content composed in Tamil and Telugu.

CUET_DUO@StressIdent_LT-EDI@EACL2024: Stress Identification Using Tamil-Telugu BERT

NLP_CUET at BLP-2023 Task 1: Fine-grained Categorization of Violence Inciting Text using Transformer-based Approach

*Coffee Breaks will be provided between 10:30 and 11:00 and 15:30 and 16:00.* To view the workshop program in detail, click [here](https://arabicnlp2024.sigarab.org) [![](https://assets.underline.io/markdown_image/1/image/12613369ab2f6dd75393ffa8794ccdd3.png)](https://app.gather.town/app/7Rf2IGR6LODvPyOx/ACL2024Workshops?spawnToken=Lcj5Wbq4QE2hykuB5Jg0)

W18: The Second Arabic Natural Language Processing Conference (ArabicNLP)

workshop paper

### Welcome!
The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. Our Virtual Poster Sessions will take place online Thursday, August 22, 2024.

You are required to register for this event. **Please register [here](https://2024.aclweb.org/registration). **

If you have already registered, please check your inbox for an email from Underline granting you access to ACL 2024 content.

Please register!

ACL 2024

The 62nd Annual Meeting of the Association for Computational Linguistics (ACL 2024) will take place in Bangkok, Thailand from August 11th to 16th, 2024. More information will be announced soon.

[**Teleport to Gather**](https://app.gather.town/app/XDzSQW1b0mpTTAYr/EACL2024?spawnToken=XxkeWnX_QkG6G-llIfSi) The development of technology increases our internet use, and most of the global languages have adapted themselves to the digital era. However, there are many regional, under-resourced languages that face challenges as they still lack developments in language technology [1]. One such language family is the Dravidian family of languages. Dravidian languages are primarily spoken in south India and Sri Lanka. Pockets of speakers are found in Nepal, Pakistan, Malaysia, other parts of India and elsewhere in the world. The Dravidian languages, which are 4,500 years old [2] and spoken by millions of speakers, are under-resourced in speech and natural language processing [1]. The Dravidian languages are divided into four groups: South, South-Central, Central, and North groups. Dravidian morphology is agglutinating and exclusively suffixal. Syntactically, Dravidian languages are head-final and left-branching. They are free-constituent order languages. To improve access to and production of information for monolingual speakers of Dravidian languages, it is necessary to have speech and languages technologies. The aim of these workshops is to save the Dravidian languages from extinction in technology. This is the first workshop on speech and language technologies for Dravidian languages. 
 [**Read more**](https://sites.google.com/view/dravidianlangtech-2024/)
 **Program Overview** 
09:15 - 09:30 Opening Remarks 
09:30 - 10:00 Keynote 
10:00 - 11:00 Developing Models and Resources for Low Resource Applications - Chair: Bharathi Raja Chakravarthi, University of Galway, Ireland 
11:00 - 13:00 Poster Session - Chair: Prasanna Kumar Kumerasan, University of Galway, Ireland 
13:00 - 13:30 Lunch Break 
13:30 - 15:45 Overview and Findings of Shared Tasks at DravidianLangTech 2024 - Chair: Rahul Ponnusamy, University of Galway, Ireland 
15:45 - 17:00 Tea Break 
17:00 - 17:15 Meeting, Awards, Closing [**Detailed Program**](https://sites.google.com/view/dravidianlangtech-2024/schedule)

W19: The Fourth Workshop on Speech and Language Technologies for Dravidian Languages (DravidianLangTech)

Welcome to the 18th Conference of the European Chapter of the Association for Computational Linguistics (EACL). Continuing its mission of expanding and involving the science community of all European countries, EACL has selected the Malta community for the 18th EACL. Considering the importance of physical interaction among researchers, the conference will be held at the Hotel Radisson Blu, St. Julians, in Malta, from 17 to 22 of March, 2024. The conference will also feature the possibility to participate virtually. As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language. 
#### [Conference Handbook](https://drive.google.com/file/d/13Wn38q0yev6U3RTgnf4Tn5T9GuAuZ-xW/view?usp=sharing) *downloadable*

To access the EACL 2024 event page you need to register [**here**](https://acl.swoogo.com/EACL2024). 

EACL 2024

As the flagship European conference in the field of computational linguistics, EACL welcomes European and international researchers covering a broad spectrum of research areas that are concerned with computational approaches to natural language.

[**Teleport to Gather**](https://app.gather.town/app/XDzSQW1b0mpTTAYr/EACL2024?spawnToken=QvCZ0RrQRM6hvHqVsTqW) Equality, Diversity and Inclusion (EDI) is an important agenda across every field throughout the world. Language as a major part of communication should be inclusive and treat everyone with equality. Today’s large internet community uses language technology (LT) and has a direct impact on people across the globe. EDI is crucial to ensure everyone is valued and included, so it is necessary to build LT that serves this purpose. Recent results have shown that big data and deep learning are entrenching existing biases and that some algorithms are even naturally biased due to problems such as ‘regression to the mode’. Our focus is on creating LT that will be more inclusive of gender, racial, sexual orientation, persons with disability. The workshop will focus on creating speech and language technology to address EDI not only in English, but also in less resourced languages. [**Read more**](https://sites.google.com/view/lt-edi-2024) [**Invited Speaker**](https://sites.google.com/view/lt-edi-2024/home) **Program Overview** 
09:00 - 09:15 Opening Remarks 
09:15 - 09:45 Keynote - Chair: Bharathi Raja Chakravarthi, University of Galway, Ireland 
09:45 - 11:00 Paper Session 1 - Chair: Bharathi Raja Chakravarthi, University of Galway, Ireland 
11:00 - 11:30 Cofee Break 
11:30 - 12:30 Poster Session 1 - Chair: B. Bharathi, SSN College of Engineering, India 
12:30 - 13:30 Lunch Break 
13:30 - 15:45 Paper Session 2 - Chair: Miguel Angel Garcia Cumbreras, University of Jaén, Spain 
15:45 - 16:15 Coffee Break 
16:15 - 17:15 Poster Session 2 - Chair: Gyorgy Kovacs,Lulea University of Technology (LTU), Sweden 
17:15 - 17:45 Plenary and Closing Remarks [**Detailed Program**](https://sites.google.com/view/lt-edi-2024/schedule)

W3: Fourth Workshop on Language Technology for Equality, Diversity and Inclusion (LTEDI)

**Workshop Organizers:** Firoj Alam, Sudipta Kar, Shammur Absar Chowdhury, Farig Sadeque and Mohammad Ruhul Amin
 *Coffee Break 10:30 am* 
*Lunch Break 12:30 pm* 
*Coffee Break 3:30 pm* [**Workshop Website**](https://blp-workshop.github.io/) This workshop aims to provide a forum for researchers to share and discuss their ongoing work with the international community. Following the success of prior local editions of the conferences in 2018 and 2019, in this first edition of the workshop, we will focus on Bangla, which is a low-resource language, and assess its current state-of-the-art and discuss strategies to make further progress in both NLP, Speech and multimodal research. Through this workshop, we plan to bring researchers together to come up with frameworks and strategies that can later support to other low-resource languages. This workshop is timely given the continued rise in research projects focusing on low-resource and multilingual studies. We particularly encourage researchers to submit their papers focusing on novel methodologies and resources that help towards the progress of Bangla and other low-resource languages. Novel methodologies include, but are not limited to, zero-shot learning, unsupervised learning, and simple yet effective methods applicable to low-computation scenarios.
 [**Go to Gather**](https://app.gather.town/app/ROH1myZvp9ba7a3C/EMNLP2023?spawnToken=e0sL3UWeS3q-7dyC3sct)

W20: The First Bangla Language Processing Workshop (BLP)

**Welcome to EMNLP 2023!**

On behalf of the EMNLP 2023 Organizing Committee, I extend a warm and heartfelt welcome to all of you to the 2023 Conference on Empirical Methods in Natural Language Processing (EMNLP). It is with immense pleasure and excitement that we gather here in Singapore, a vibrant hub of innovation and technological advancement.

The conference program is packed with insightful presentations, thought-provoking workshops, and engaging networking opportunities. In addition to the technical sessions, we have also planned several social events that will provide you with the opportunity to connect with your colleagues.

The conference is held in person at the Resorts World Convention Centre in Singapore, and available on-line with the help of Underline.

For those in Singapore, I hope that you will find time to explore the exciting Sentosa Island, Gardens by the Bay, Singapore Botanic Gardens and the many other attractions unique to Singapore.

Of course, we are grateful to our sponsors and partners for their generous support of EMNLP 2023, whose contributions make it possible for us to host this world-class event.

Yuji Matsumoto (RIKEN AIP) 
EMNLP 2023 General Chair

To access the **EMNLP 2023** event page on Underline, you need to register for the Conference. 
Please follow **[this link](https://2023.emnlp.org/registration/)** for more details.

Jawad Hossain

15

8

SHORT BIO

Presentations