PrefPalette: Personalized Preference Modeling with Latent Attributes

TL;DR

PrefPalette is a novel framework that decomposes user preferences into interpretable attributes and models community-specific values, significantly improving prediction accuracy and providing transparent insights into human judgment.

Contribution

It introduces a scalable method for preference prediction that incorporates attribute effects and social community values, advancing personalized AI with interpretability.

Findings

Outperforms GPT-4o by 46.6% in accuracy on Reddit communities.

Reveals community-specific preference profiles like verbosity, sarcasm, and empathy.

Enhances understanding of human judgment through attribute-based modeling.

Abstract

Personalizing AI systems requires understanding not just what users prefer, but the reasons that underlie those preferences - yet current preference models typically treat human judgment as a black box. We introduce PrefPalette, a framework that decomposes preferences into attribute dimensions and tailors its preference prediction to distinct social community values in a human-interpretable manner. PrefPalette operationalizes a cognitive science principle known as multi-attribute decision making in two ways: (1) a scalable counterfactual attribute synthesis step that involves generating synthetic training data to isolate for individual attribute effects (e.g., formality, humor, cultural values), and (2) attention-based preference modeling that learns how different social communities dynamically weight these attributes. This approach moves beyond aggregate preference modeling to capture the diverse evaluation frameworks that drive human judgment. When evaluated on 45 social communities from the online platform Reddit, PrefPalette outperforms GPT-4o by 46.6% in average prediction accuracy. Beyond raw predictive improvements, PrefPalette also shed light on intuitive, community-specific profiles: scholarly communities prioritize verbosity and stimulation, conflict-oriented communities value sarcasm and directness, and support-based communities emphasize empathy. By modeling the attribute-mediated structure of human judgment, PrefPalette delivers both superior preference modeling and transparent, interpretable insights, and serves as a first step toward more trustworthy, value-aware personalized applications.