Do You Know My Emotion? Emotion-Aware Strategy Recognition towards a Persuasive Dialogue System

TL;DR

This paper introduces CFO-Net, a novel emotion-aware model for persuasive strategy recognition that leverages emotional feedback to enhance understanding of dialogue strategies, showing improved performance on a persuasion dataset.

Contribution

The paper proposes CFO-Net, which incorporates emotional feedback into strategy recognition using a feedback memory module and cross-channel fusion, advancing dialogue system capabilities.

Findings

CFO-Net improves M-F1 score from 61.74 to 65.41 on PersuasionForGood.

Emotion-aware feedback enhances strategy recognition accuracy.

The model effectively integrates emotional cues into dialogue understanding.

Abstract

Persuasive strategy recognition task requires the system to recognize the adopted strategy of the persuader according to the conversation. However, previous methods mainly focus on the contextual information, little is known about incorporating the psychological feedback, i.e. emotion of the persuadee, to predict the strategy. In this paper, we propose a Cross-channel Feedback memOry Network (CFO-Net) to leverage the emotional feedback to iteratively measure the potential benefits of strategies and incorporate them into the contextual-aware dialogue information. Specifically, CFO-Net designs a feedback memory module, including strategy pool and feedback pool, to obtain emotion-aware strategy representation. The strategy pool aims to store historical strategies and the feedback pool is to obtain updated strategy weight based on feedback emotional information. Furthermore, a cross-channel fusion predictor is developed to make a mutual interaction between the emotion-aware strategy representation and the contextual-aware dialogue information for strategy recognition. Experimental results on \textsc{PersuasionForGood} confirm that the proposed model CFO-Net is effective to improve the performance on M-F1 from 61.74 to 65.41.