Distinguishing fact from opinion in text is a nuanced but essential task, particularly in news articles where subjectivity can influence interpretation and reception. Identifying whether content is subjective or objective is critical for sentiment analysis, media bias detection, and content moderation. However, progress in this area has been limited for low-resource languages such as Bengali due to a lack of benchmark datasets and tools. To address these constraints, this work presents BeNSD (Bengali News Subjectivity Detection), a novel dataset of 8,655 Bengali news article texts, along with an enhanced transformer-based architecture (POS-Aware-MuRIL) that integrates parts-of-speech (POS) features with MuRIL embeddings at the input level to provide richer contextual representation for subjectivity detection. A range of baseline models is evaluated, and the proposed architecture achieves a macro F1-score of 93.35% in subjectivity detection for the Bengali language.

Hate speech on social media platforms, particularly in low-resource languages like Bengali, poses a significant challenge due to its nuanced nature and the need to understand its type, severity, and targeted group. To address this, the Bangla Multi-task Hate Speech Identification Shared Task at BLP 2025 adopts a multi-task learning framework that requires systems to classify Bangla YouTube comments across three subtasks simultaneously: type of hate, severity, and targeted group. To tackle these challenges, this work presents BanTriX, a transformer ensemble method that leverages BanglaBERT-I, XLM-R, and BanglaBERT-II. Evaluation results show that the BanTriX, optimized with cross-entropy loss, achieves the highest weighted micro F1-score of 73.78% in Subtask 1C, securing our team 2nd place in the shared task.

Detecting hate speech in Bengali social media content presents considerable challenges, primarily due to the prevalence of informal language and the limited availability of annotated datasets. This study investigates the identification of hate speech in Bengali YouTube comments, focusing on classifying the type, severity, and target group. Multiple machine learning baselines and voting ensemble techniques are evaluated to address these tasks. The methodology involves text preprocessing, feature extraction using TF-IDF and Count vectors, and aggregating predictions from several models. Hierarchical classification with TF-IDF features and majority voting improves the detection of less frequent hate speech categories while maintaining robust overall performance, resulting in an 18^th place ranking and a micro F1 score of 68.42%. Furthermore, ablation studies assess the impact of preprocessing steps and n-gram selection, providing reproducible baselines for Bengali hate speech detection. All codes and resources are publicly available at https://github.com/Hasan-Mesbaul-Ali-Taher/BLP_25_Task_1

Fake news detection is a critical problem in today’s digital age, aiming to classify intentionally misleading or fabricated news content. In this study, we present a transformer and LLM-based ensemble method to address the challenges in fake news detection. We explored various machine learning (ML), deep learning (DL), transformer, and LLM-based approaches on a Malayalam fake news detection dataset. Our findings highlight the difficulties faced by traditional ML and DL methods in accurately detecting fake news, while transformer- and LLM-based ensemble methods demonstrate significant improvements in performance. The ensemble method combining Sarvam-1, Malayalam-BERT, and XLM-R outperformed all other approaches, achieving an F1-score of 89.30% on the given dataset. This accomplishment, which contributed to securing 2nd place in the shared task at DravidianLangTech 2025, underscores the importance of developing effective methods for detecting fake news in Dravidian languages.

Misogyny memes, a form of digital content, reflect societal prejudices by discriminating against women through shaming and stereotyping. In this study, we present a multimodal approach combining Indic-BERT and ViT-base-patch16-224 to address misogyny memes. We explored various Machine Learning, Deep Learning, and Transformer models for unimodal and multimodal classification using provided Tamil and Malayalam meme dataset. Our findings highlight the challenges traditional ML and DL models face in understanding the nuances of Dravidian languages, while emphasizing the importance of transformer models in capturing these complexities. Our multimodal method achieved F1-scores of 77.18% and 84.11% in Tamil and Malayalam, respectively, securing 6th place for both languages among the participants.

Large Language Models (LLMs) demonstrate exceptional proficiency in general-purpose tasks but struggle with numerical reasoning, particularly in low-resource languages like Bengali. Despite advancements, limited research has explored their numerical reasoning capabilities in these languages. To address this gap, we present BenNumEval (Bengali Numerical Evaluation), a benchmark designed to assess LLMs on numerical reasoning tasks in Bengali. It comprises six diverse tasks and a total of 3.2k samples curated from real-world problem-solving scenarios. Our extensive evaluations reveal that even with advanced prompting techniques such as Cross-Lingual Prompting (XLP) and Cross-Lingual Chain-of-Thought Prompting (XCoT), LLMs fall notably short of human-level performance, particularly when using Bengali Native Prompting (BNaP). These findings underscore the substantial gap between current LLM capabilities and human expertise in numerical reasoning, highlighting the need for more robust and linguistically inclusive AI models to advance Bengali Language Processing and equitable AI development. The source code for the system and evaluation pipeline is publicly available on GitHub.

Memes are now a common means of communication on social media. Their humor and short format help messages spread quickly and easily. Propagandistic memes use both words and images to influence opinions and behaviors, often appealing to emotions or ideologies. While propaganda detection has been well-studied in high-resource languages (HRLs), there has been a limited focus on low-resource languages (LRLs), such as Bengali. In this study, we introduce MemeGuard, a new dataset of 3,745 memes for detecting propaganda in Bengali. We tested more than 45 different methods, including both single and combined approaches with fusion. For text, BanglaBERT-1 achieved the best macro F1 score of 80.34%, whereas the CLIP vision transformer scored 78.94% for images. The proposed multimodal model, which combines BanglaBERT-2 and CLIP using Adaptive Modality Fusion, achieved the highest macro F1 score of 85.36%. This work establishes a strong baseline and offers valuable insights for future research in Bengali multimodal content analysis.

In this modern era, many people have been using Facebook and Twitter, leading to increased 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 as fake news. A shared task on fake news detection in Malayalam organized by DravidianLangTech@EACL 2024 allowed us for addressing the challenge of distinguishing between original and fake news content in the Malayalam language. Our approach involves creating an intelligent framework to categorize text as either fake or original. We experimented with various machine learning models, including Logistic Regression, Decision Tree, Random Forest, Multinomial Naive Bayes, SVM, and SGD, and various deep learning models, including CNN, BiLSTM, and BiLSTM + Attention. We also explored Indic-BERT, MuRIL, XLM-R, and m-BERT for transformer-based approaches. Notably, our most successful model, m-BERT, achieved a macro F1 score of 0.85 and ranked 4th in the shared task. This research contributes to combating misinformation on social media news, offering an effective solution to classify content accurately.

Violence-inciting text detection has become critical due to its significance in social media monitoring, online security, and the prevention of violent content. Developing an automatic text classification model for identifying violence in languages with limited resources, like Bangla, poses significant challenges due to the scarcity of resources and complex morphological structures. This work presents a transformer-based method that can classify Bangla texts into three violence classes: direct, passive, and non-violence. We leveraged transformer models, including BanglaBERT, XLM-R, and m-BERT, to develop a hierarchical classification model for the downstream task. In the first step, the BanglaBERT is employed to identify the presence of violence in the text. In the next step, the model classifies stem texts that incite violence as either direct or passive. The developed system scored 72.37 and ranked 14th among the participants.