Exploring Persona Sentiment Sensitivity in Personalized Dialogue Generation

TL;DR

This paper investigates how the sentiment polarity of user personas affects the quality and characteristics of dialogues generated by large language models, revealing that sentiment influences persona emphasis and interaction smoothness.

Contribution

It provides a large-scale analysis of persona sentiment effects on dialogue quality and proposes a sentiment-aware generation method to improve personalized responses.

Findings

Negatively polarized users lead to overemphasized persona attributes.

Positively polarized profiles produce smoother, more balanced dialogues.

Neutral or weak sentiment personas generate lower-quality interactions.

Abstract

Personalized dialogue systems have advanced considerably with the integration of user-specific personas into large language models (LLMs). However, while LLMs can effectively generate personalized responses, the influence of persona sentiment on dialogue quality remains underexplored. In this work, we conduct a large-scale analysis of dialogues generated using a range of polarized user profiles. Our experiments reveal that dialogues involving negatively polarized users tend to overemphasize persona attributes. In contrast, positively polarized profiles yield dialogues that selectively incorporate persona information, resulting in smoother interactions. Furthermore, we find that personas with weak or neutral sentiment generally produce lower-quality dialogues. Motivated by these findings, we propose a dialogue generation approach that explicitly accounts for persona polarity by combining a turn-based generation strategy with a profile ordering mechanism and sentiment-aware prompting. Our study provides new insights into the sensitivity of LLMs to persona sentiment and offers guidance for developing more robust and nuanced personalized dialogue systems.