How Stylistic Similarity Shapes Preferences in Dialogue Dataset with User and Third Party Evaluations

TL;DR

This paper introduces a new dataset to analyze how subjective and objective stylistic similarities influence user preferences in dialogue systems, revealing that subjective perceptions strongly correlate with user satisfaction and differ from third-party evaluations.

Contribution

The paper presents a novel dataset with subjective and objective stylistic similarity annotations, highlighting the importance of distinguishing these measures in dialogue preference analysis.

Findings

Strong correlation between subjective stylistic similarity and user preference

Users' subjective similarity differs from third-party objective similarity

The dataset enables better understanding of stylistic factors in dialogue preferences

Abstract

Recent advancements in dialogue generation have broadened the scope of human-bot interactions, enabling not only contextually appropriate responses but also the analysis of human affect and sensitivity. While prior work has suggested that stylistic similarity between user and system may enhance user impressions, the distinction between subjective and objective similarity is often overlooked. To investigate this issue, we introduce a novel dataset that includes users' preferences, subjective stylistic similarity based on users' own perceptions, and objective stylistic similarity annotated by third party evaluators in open-domain dialogue settings. Analysis using the constructed dataset reveals a strong positive correlation between subjective stylistic similarity and user preference. Furthermore, our analysis suggests an important finding: users' subjective stylistic similarity differs from third party objective similarity. This underscores the importance of distinguishing between subjective and objective evaluations and understanding the distinct aspects each captures when analyzing the relationship between stylistic similarity and user preferences. The dataset presented in this paper is available online.