The rapid adoption of AI in educational technology is changing learning settings, making the thorough evaluation of AI tutor pedagogical performance is quite important for promoting student success. This paper describes our solution for the BEA 2025 Shared Task on Pedagogical Ability Assessment of AI-powered tutors, which assesses tutor replies over several pedagogical dimensions. We developed transformer-based approaches for five diverse tracks: mistake identification, mistake location, providing guidance, actionability, and tutor identity prediction using the MRBench dataset of mathematical dialogues. We evaluated several pre-trained models including DeBERTa-V3, RoBERTa-Large, SciBERT, and EduBERT. Our approach addressed class imbalance problems by incorporating strategic fine-tuning with weighted loss functions. The findings show that, for all tracks, DeBERTa architectures have higher performances than the others, and our models have obtained in the competitive positions, including 9th of Tutor Identity (Exact F1 of 0.8621), 16th of Actionability (Exact F1 of 0.6284), 19th of Providing Guidance (Exact F1 of 0.4933), 20th of Mistake Identification (Exact F1 of 0.6617) and 22nd of Mistake Location (Exact F1 of 0.4935). The difference in performance over tracks highlights the difficulty of automatic pedagogical evaluation, especially for tasks whose solutions require a deep understanding of educational contexts. This work contributes to ongoing efforts to develop robust automated tools for assessing.

E-commerce platforms face growing challenges regarding consumer trust and review authenticity because of the growing number of AI-generated product reviews. Low-resource languages (LRLs) such as Tamil and Malayalam face limited investigation by AI detection techniques because these languages experience constraints from sparse data sources and complex linguistic structures. The research team at CUET_NetworkSociety took part in the AI-Generated Review Detection contest during the DravidianLangTech@NAACL 2025 event to fill this knowledge void. Using a combination of machine learning, deep learning, and transformer-based models, we detected AI-generated and human-written reviews in both Tamil and Malayalam. The developed method employed DistilBERT, which underwent an advanced preprocessing pipeline and hyperparameter optimization using the Transformers library. This approach achieved a Macro F1-score of 0.81 for Tamil (Subtask 1), securing 18th place, and a score of 0.7287 for Malayalam (Subtask 2), ranking 25th.

Memes are commonly used for communication on social media platforms, and some of them can propagate misogynistic content, spreading harmful messages. Detecting such misogynistic memes has become a significant challenge, especially for low-resource languages like Tamil and Malayalam, due to their complex linguistic structures. To tackle this issue, a shared task on detecting misogynistic memes was organized at DravidianLangTech@NAACL 2025. This paper proposes a multimodal deep learning approach for detecting misogynistic memes in Tamil and Malayalam. The proposed model combines fine-tuned ResNet18 for visual feature extraction and indicBERT for analyzing textual content. The fused model was applied to the test dataset, achieving macro F1 scores of 76.32% for Tamil and 80.35% for Malayalam. Our approach led to 7th and 12th positions for Tamil and Malayalam, respectively.

Social media has become an established medium of public communication and opinions on every aspect of life, but especially politics. This has resulted in a growing need for tools that can process the large amount of unstructured data that is produced on these platforms providing actionable insights in domains such as social trends and political opinion. Low-resource languages like Tamil present challenges due to limited tools and annotated data, highlighting the need for NLP focus on understudied languages. To address this, a shared task has been organized by DravidianLangTech@NAACL 2025 for political sentiment analysis for low-resource languages, with a specific focus on Tamil. In this task, we have explored several machine learning methods such as SVM, AdaBoost, GB, deep learning methods including CNN, LSTM, GRU BiLSTM, and the ensemble of different deep learning models, and transformer-based methods including mBERT, T5, XLM-R. The mBERT model performed best by achieving a macro F1 score of 0.2178 and placing our team 22nd in the rank list.

This paper presents our system for the detection of racial hoaxes in code-mixed Hindi-English social media narratives, which is in reality a form of debunking of online disinformation claiming fake incidents against a racial group. We experiment with different modeling techniques on HoaxMixPlus dataset of 5,102 annotated YouTube comments. In our approach, we utilize traditional machine learning classifiers (SVM, LR, RF), deep learning models (CNN, CNN-LSTM, CNN-BiLSTM), and transformer-based architectures (MuRIL, XLM-RoBERTa, HingRoBERTa-mixed). Experiments show that transformer-based methods substantially outperform traditional approaches, and the HingRoBERTa-mixed model is the best one with an F1 score of 0.7505. An error analysis identifies the difficulty of recognizing implicit bias and nuanced contexts in complex hoaxes. Our team was 5th place in the challenge with an F1 score of 0.69. This work contributes to combating online misinformation in low-resource linguistic environments and highlights the effectiveness of specialized language models for code-mixed content.

This paper presents our system for detecting caste and migration-related hate speech in Tamil social media comments, addressing the challenges in this low-resource language setting. We experimented with multiple approaches on a dataset of 7,875 annotated comments. Our methodology encompasses traditional machine learning classifiers (SVM, Random Forest, KNN), deep learning models (CNN, CNN-BiLSTM), and transformer-based architectures (MuRIL, IndicBERT, XLM-RoBERTa). Comprehensive evaluations demonstrate that transformer-based models substantially outperform traditional approaches, with MuRIL-large achieving the highest performance with a macro F1 score of 0.8092. Error analysis reveals challenges in detecting implicit and culturally-specific hate speech expressions requiring deeper socio-cultural context. Our team ranked 5th in the LT-EDI@LDK 2025 shared task with an F1 score of 0.80916. This work contributes to combating harmful online content in low-resource languages and highlights the effectiveness of large pre-trained multilingual models for nuanced text classification tasks.

Language is a rich medium employed to convey emotions subtly and intricately, as abundant as human emotional experiences themselves. Emotion recognition in natural language processing (NLP) is now a core element in facilitating human-computer interaction and interpreting intricate human behavior via text. It has potential applications in every sector i.e., sentiment analysis, mental health surveillance. However, prior research on emotion recognition is primarily from high-resource languages while low-resource languages (LRLs) are not well represented. This disparity has been a limitation to the development of universally applicable emotion detection models. To address this, the SemEval-2025 Shared Task 11 focused on perceived emotions, aiming to identify the emotions conveyed by a text snippet. It includes three tracks: Multi-label Emotion Detection (Track A), Emotion Intensity (Track B), and Cross-lingual Emotion Detection (Track C). This paper explores various models, including machine learning (LR, SVM, RF, NB), deep learning (BiLSTM+CNN, BiLSTM+BiGRU), and transformer-based models (XLM-R, mBERT, ModernBERT). The results showed that XLM-R outperformed other models in Tracks A and B, while BiLSTM+CNN performed better for Track C across most languages.