Organizations: Department of Educational Technology and Engineering, University of Frontier Technology, Bangladesh (UFTB), Gazipur, Bangladesh. · Department of Data Science and Engineering, University of Frontier Technology, Bangladesh (UFTB), Gazipur, Bangladesh. · Department of IOT and Robotics Engineering, University of Frontier Technology, Bangladesh (UFTB), Gazipur, Bangladesh.
The recent escalation of the Iran Israel USA conflict in 2026 has triggered widespread global discussions across social media platforms. As people increasingly use these platforms for expressing opinions, analyzing public sentiment from these discussions can provide valuable insights into global public perception. This study aims to analyze global public sentiment regarding the Iran Israel USA conflict by mining user-generated comments from YouTube news channels. The work contributes to public opinion analysis by introducing a privacy preserving framework that combines topic wise sentiment analysis with modern deep learning techniques and Federated Learning. To achieve this, approximately 19,000 YouTube comments were collected from major international news channels and preprocessed to remove noise and normalize text. Sentiment labels were initially generated using the VADER sentiment analyzer and later validated through manual inspection to improve reliability. Latent Dirichlet Allocation (LDA) was applied to identify key discussion topics related to the conflict. Several transformer-based models, including BERT, RoBERTa, XLNet, DistilBERT, ModernBERT, and ELECTRA, were fine tuned for sentiment classification. The best-performing model was further integrated into a federated learning environment to enable distributed training by preserving user data privacy. Additionally, Explainable Artificial Intelligence (XAI) techniques using SHAP were applied to interpret model predictions and identify influential words affecting sentiment classification. Experimental results demonstrate that transformer models perform effectively, and among them, ELECTRA achieved the best performance with 91.32% accuracy. The federated learning also maintained strong performance while preserving privacy, achieving 89.59% accuracy in a two client configuration.
Figures & tables
Figure 1: Overview of the data preprocessing pipeline
Figure 2: Illustration of the ELECTRA generator and discriminator framework.
Figure 3: Architecture of the proposed Federated ELECTRA model.
Figure 4: Overall workflow of the proposed framework including data preprocessing, sentiment labeling, topic modeling, transformer-based model comparison, federated learning and explainable AI analysis.
Figure 5: Topic distribution of USA–Iran war related comments generated using LDA.
Figure 6: Sentiment distribution across different topics showing the number of negative, neutral, and positive comments.
Figure 7: Word cloud visualization representing the most frequent words appearing in positive sentiment comments.
Figure 8: Word cloud visualization showing the most common words appearing in negative sentiment comments.
Figure 9: Word cloud visualization of frequently used words in neutral sentiment comments.
Model
Accuracy (%)
Macro F1 (%)
Weighted F1 (%)
BERT
90.68
90.43
90.67
RoBERTa
90.23
89.97
90.22
XLNet
90.73
90.43
90.72
DistilBERT
90.37
90.22
90.36
ELECTRA
91.32
90.96
91.29
ModernBERT
90.84
90.55
90.84
Table 1: Comparative performance of transformer-based models for sentiment classification
Figure 10: Confusion matrix of the ELECTRA-based sentiment classification model.
Number of Clients
Accuracy (%)
Macro F1 (%)
Weighted F1 (%)
2 Clients
89.59
89.30
89.57
4 Clients
86.88
86.23
86.79
6 Clients
85.27
84.29
85.14
Table 2: Performance of the ELECTRA-based federated learning model under different client configurations
Figure 11: Confusion matrixes of the ELECTRA-based federated learning model under different client configurations: (a) 2 clients, (b) 4 clients, and (c) 6 clients. The matrices illustrate the distribution of predicted versus true sentiment labels across the negative, neutral, and positive classes.
Figure 12: Class-wise SHAP explanation showing the most influential tokens contributing to negative, neutral, and positive sentiment predictions. Red bars indicate tokens that increase the likelihood of the class, while blue bars represent tokens that decrease the probability of that class.
YouTube Shorts have become central to news consumption on the platform, yet research on how geopolitical events are represented in this format remains limited. To address this gap, we present a multimodal pipeline that combines automatic transcription, aspect-based sentiment analysis (ABSA), and semantic scene classification. The pipeline is first assessed for feasibility and then applied to analyze short-form coverage of the Israel-Hamas war by state-funded outlets. Using over 2,300 conflict-related Shorts and more than 94,000 visual frames, we systematically examine war reporting across major international broadcasters. Our findings reveal that the sentiment expressed in transcripts regarding specific aspects differs across outlets and over time, whereas scene-type classifications reflect visual cues consistent with real-world events. Notably, smaller domain-adapted models outperform large transformers and even LLMs for sentiment analysis, underscoring the value of resource-efficient approaches for humanities research. The pipeline serves as a template for other short-form platforms, such as TikTok and Instagram, and demonstrates how multimodal methods, combined with qualitative interpretation, can characterize sentiment patterns and visual cues in algorithmically driven video environments.
We present StanceNakba 2026, a shared task on stance detection in polarized social media discourse related to the Palestinian-Israeli conflict, organized as part of Nakba-NLP 2026 at LREC-COLING 2026. The task introduces two subtasks: Subtask A (Actor-Level Stance Detection), which classifies English social media posts as Pro-Palestine, Pro-Israel, or Neutral; and Subtask B (Cross-Topic Stance Detection), which identifies Favor, Against, or Neither stances in Arabic posts toward two conflict-related topics, normalization with Israel and refugee presence in Jordan. The task is grounded in an annotated dataset of 2,606 social media posts. A total of 7 teams participated in Subtask A and 6 teams in Subtask B. Participating systems primarily fine-tuned Arabic and multilingual transformer-based models, including MARBERT, AraBERT, and DeBERTa-v3 variants, with several teams employing cross-validation, ensemble methods, and topic-conditioned architectures. The best-performing systems achieved a Macro F1 of 0.9620 on Subtask A and 0.8724 on Subtask B, demonstrating that transformer-based approaches are highly effective for conflict-domain stance detection while highlighting persistent challenges in cross-topic generalization and neutral class prediction.
Kholoud K. Aldous, Md Rafiul Biswas, Mabrouka Bessghaier +3
Northwestern University in Qatar · Hamad Bin Khalifa University
This study examines how different artificial intelligence architectures interpret sentiment in conflict-related media discourse, using the 2023 Gaza War as a case study. Drawing on a corpus of 10,990 Arabic news headlines (Eleraqi 2026), the research conducts a comparative analysis between three large language models and six fine-tuned Arabic BERT models. Rather than evaluating accuracy against a single human-annotated gold standard, the study adopts an epistemological approach that treats sentiment classification as an interpretive act produced by model architectures. To quantify systematic differences across models, the analysis employs information-theoretic and distributional metrics, including Shannon Entropy, Jensen-Shannon Distance, and a Variance Score measuring deviation from aggregate model behavior. The results reveal pronounced and non-random divergence in sentiment distributions. Fine-tuned BERT models, particularly MARBERT, exhibit a strong bias toward neutral classifications, while LLMs consistently amplify negative sentiment, with LLaMA-3.1-8B showing near-total collapse into negativity. Frame-conditioned analysis further demonstrates that GPT-4.1 adjusts sentiment judgments in line with narrative frames (e.g., humanitarian, legal, security), whereas other LLMs display limited contextual modulation. These findings suggest that the choice of model constitutes a choice of interpretive lens, shaping how conflict narratives are algorithmically framed and emotionally evaluated. The study contributes to media studies and computational social science by foregrounding algorithmic discrepancy as an object of analysis and by highlighting the risks of treating automated sentiment outputs as neutral or interchangeable measures of media tone in contexts of war and crisis.
Amr Eleraqi, Hager H. Mustafa, Abdul Hadi N. Ahmed
Faculty of Graduate Studies and Research, Cairo University, Egypt · Independent Researcher, Anmat Media