Rewarding Sentiment Consistency: Reinforcement Learning for Multilingual Summarization

doi:10.1007/978-3-032-09044-7_2

Rewarding Sentiment Consistency: Reinforcement Learning for Multilingual Summarization

2026-01-01

0

PRE

AI

Liliana Chanona-Hernández

DOI：10.1007/978-3-032-09044-7_2

Original

Original request for help

Save

Abstract

En 中文

This study addresses sentiment preservation in multilingual text summarization by proposing a reinforcement learning framework with a composite reward function. The approach combines sentiment consistency, content retention, fluency, and diversity metrics, leveraging an adapted mT5 model with language-specific processing. Evaluation across five languages shows 12-15% improvement in sentiment preservation, with particularly strong results (34% reduction in errors) for morphologically complex languages like Finnish. The framework demonstrates practical value in e-commerce and mental health applications, achieving 22 higher customer satisfaction and 18% better emotional tone detection. While effective, computational overhead and challenges with low-resource languages remain for future optimization.

Keywords:

Multilingual summarization

Sentiment preservation

Reinforcement learning

Morphological complexity

Rewarding Sentiment Consistency: Reinforcement Learning for Multilingual Summarization

Abstract

Journal

Organization