Agent guided evidence improves emotion cause extraction accuracy
Agent-Guided Evidence Acquisition for Emotion-Cause Pair Extraction
Computation and Language
Summary
Finding out what causes emotions in conversations can be tricky when all the information isn’t available right away. This paper studies a smart helper that asks experts only when it’s unsure, focusing on the most uncertain cases to improve its guesses. The helper changes its answers only based on the extra expert advice it gets, which leads to better results without needing too much extra information. The authors showed that being selective about when and what to ask can make the system more accurate and efficient.
What this means in practice
- •For conversational ai developers: Improve chatbot understanding of emotional context by selectively consulting external knowledge only for uncertain emotional cause predictions.
- •For customer service automation teams: Enhance emotion cause detection accuracy in automated sentiment analysis tools by integrating targeted expert feedback on ambiguous cases.
Authors
Zhuangzhuang Pan, Yan Xia, Chee Seng Chan
Abstract
Emotion-Cause Pair Extraction (ECPE) requires identifying the causes associated with emotions expressed in a conversation. Accurate extraction can depend on evidence that is not initially available, making it important to decide when further information is worth acquiring. We study agent-guided evidence acquisition for ECPE, where a student agent selectively accesses expert judgments for borderline relations on both sides of its validation-selected F1 decision boundary and revises only queried predictions. Through controlled comparisons under different information-access conditions, we examine how delegation granularity, query selection, and response integration affect prediction quality. At a nominal 10% query rate, selective fusion improves mean pair F1 by 2.64-3.88 points over the student in our main experiments, reaching 65.14 on ECF and 49.74 on MEC4 across model families. Our analysis suggests that greater access does not always translate into better cause identification. These findings motivate acquisition strategies that consider both the relevance of newly available evidence and its effect on prediction revision, providing an empirical basis for more effective and efficient evidence acquisition in ECPE.