A reinforcement-driven multiple-instance learning method for multi-task speaker attribute prediction
Iran Journal of Computer Science · 2026-12 · 9 (1)
DOI: 10.1007/s42044-026-00407-3 ↗Kurz gefasst
Fachartikel in der Zeitschrift Iran Journal of Computer Science, online erschienen am 18.08.2026. Verfasst von Stijn Lakeman und 2 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Standard models for predicting speaker attributes from text often fail to manage multiple, interdependent attributes simultaneously, and existing Reinforced Multiple-Instance Learning (RL–MIL) frameworks are typically limited to single-task prediction. To address this, we propose and evaluate a Reinforced, Multi-Task, Multiple-Instance Learning framework, a novel framework that enhances Reinforced Multiple-Instance Learning with Multi-Task Learning (MTL) to predict a speaker’s age, gender, and political party from congressional speeches. A central goal of our work was to investigate the optimal parameter-sharing strategy. We compared a fully shared architecture against a dynamic task clustering mechanism designed to mitigate negative transfer by adaptively grouping related tasks. Our results demonstrate that the multi-task approach significantly outperforms single-task baselines. Interestingly, the model with a fully shared representation achieved the highest macro-average F1 score of 0.668, suggesting the tasks in this dataset were sufficiently correlated to benefit from shared features without needing adaptive separation. This work contributes a more effective method for weakly supervised, multi-attribute prediction and provides crucial insights into the trade-offs between different parameter-sharing strategies in MTL.
Zitieren
APA
Lakeman, S., Ziabari, S. S. M., & Alsahag, A. M. M. (2026). A reinforcement-driven multiple-instance learning method for multi-task speaker attribute prediction. Iran Journal of Computer Science, 9(1). https://doi.org/10.1007/s42044-026-00407-3
Vancouver
Lakeman S, Ziabari SSM, Alsahag AMM. A reinforcement-driven multiple-instance learning method for multi-task speaker attribute prediction. Iran Journal of Computer Science. 2026;9(1). doi:10.1007/s42044-026-00407-3
BibTeX
@article{Lakeman2026,
author = {Lakeman, Stijn and Ziabari, Seyed Sahand Mohammadi and Alsahag, Ali Mohammed Mansoor},
title = {A reinforcement-driven multiple-instance learning method for multi-task speaker attribute prediction},
journal = {Iran Journal of Computer Science},
year = {2026},
volume = {9},
number = {1},
doi = {10.1007/s42044-026-00407-3}
}
EUVIMED