Claim Detection From Speech in a Low-Resource Language: A New Annotated Dataset and Temporal Modeling Using Time-Series Transformer
ACM Transactions on Social Computing · 2026-11-16
DOI: 10.1145/3862801 ↗At a glance
Journal article in ACM Transactions on Social Computing, published online on 2026-11-16. Written by Eity Roy and 2 co-authors. 40 references.
Abstract (original language)
Claim Detection involves the task of recognizing claims within an utterance, differentiating between claims and non-claims. Although significant progress has been achieved in text-based claim detection for high-resource languages, audio-based claim detection remains largely unexplored, particularly in low-resource languages such as Bangla. To the best of our knowledge, our study presents the first systematic approach to audio-based claim detection for Bangla and introduces the first annotated Bangla audio claim detection dataset comprising 4,000 manually annotated speech samples, equally divided between claim and non-claim instances. We evaluate acoustic, ASR-based text, and combined multi-modal baselines under a leakage-safe, speaker- and video-disjoint evaluation protocol. We adapt a Transformer encoder with Learned pos. encoding (projected one-hot) for frame-level MFCC sequences, demonstrating its effectiveness for temporal acoustic modeling in Bangla, a low-resource language. Unlike conventional classifiers that rely on aggregated acoustic statistics, the proposed model preserves frame-level temporal structure and employs multi-head self-attention to capture long-range dependencies across speech segments. A two-phase training strategy is further adopted. Its contribution lies in demonstrating that this adaptation is effective for Bangla audio claim detection, not in architectural novelty. To provide a comprehensive benchmark, we evaluate traditional machine learning models (Logistic Regression, Random Forest, SVM, XGBoost), deep learning architectures (CNN, LSTM, CNN-LSTM hybrids, DNN, CNN+BiLSTM and CRNN+Attention), transformer-based models (XLS-R, Vision Transformer, AST, XLM-R), large self-supervised speech models (WavLM-Large, HuBERT-Large, XLS-R-300M), text SSL models (BanglaBERT, XLM-RoBERTa, mDeBERTa-V3) with parameter-efficient fine-tuning and an audio-capable large language model (Qwen2-Audio-7B). The proposed Time-Series Transformer achieves the strongest overall performance at 84.4% accuracy, performing comparably to the best text-based baseline (XLM-RoBERTa Large, 83.0%) while significantly outperforming the strongest audio-based deep learning and self-supervised baselines, and eliminating the need for an intermediate ASR transcription pipeline. The audio-capable LLM Qwen2-Audio-7B achieves only 53% accuracy despite its 7B parameter scale, confirming that model size alone does not compensate for language and domain gap in low-resource settings.
Cite
APA
Roy, E., Rahman, M. R., & Sarkar, S. (2026). Claim Detection From Speech in a Low-Resource Language: A New Annotated Dataset and Temporal Modeling Using Time-Series Transformer. ACM Transactions on Social Computing. https://doi.org/10.1145/3862801
Vancouver
Roy E, Rahman MR, Sarkar S. Claim Detection From Speech in a Low-Resource Language: A New Annotated Dataset and Temporal Modeling Using Time-Series Transformer. ACM Transactions on Social Computing. 2026. doi:10.1145/3862801
BibTeX
@article{Roy2026,
author = {Roy, Eity and Rahman, Md. Rashadur and Sarkar, Suranjan},
title = {Claim Detection From Speech in a Low-Resource Language: A New Annotated Dataset and Temporal Modeling Using Time-Series Transformer},
journal = {ACM Transactions on Social Computing},
year = {2026},
doi = {10.1145/3862801}
}
EUVIMED