Similarity = Value? Consultation Value Assessment and Alignment for Personalized Search

2025-06-17

AI Summary

Existing personalized search methods predominantly rely on semantic similarity between historical consultations and the current query, neglecting explicit modeling of consultation “value,” thereby limiting personalization efficacy. Method: This paper introduces the first value-oriented—rather than similarity-oriented—evaluation paradigm for e-commerce AI assistants. We propose a three-dimensional value framework encompassing contextual scope value, posterior behavioral value, and temporal decay value. Within a Transformer architecture, we design a consultation–user behavioral interaction module and an explicit alignment objective, integrating multi-granularity behavioral modeling, temporal weighting, contrastive learning loss, and an interpretable filtering mechanism for end-to-end value-aware ranking. Results: Evaluated on both public and real-world commercial datasets, our approach achieves +3.2% AUC and +4.7% NDCG@10 over state-of-the-art methods, comprehensively validating the effectiveness and practicality of the value-driven paradigm.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Machine Learning: Learning Preferences or Rankings
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Personalized search systems in e-commerce platforms increasingly involve user interactions with AI assistants, where users consult about products, usage scenarios, and more. Leveraging consultation to personalize search services is trending. Existing methods typically rely on semantic similarity to align historical consultations with current queries due to the absence of 'value' labels, but we observe that semantic similarity alone often fails to capture the true value of consultation for personalization. To address this, we propose a consultation value assessment framework that evaluates historical consultations from three novel perspectives: (1) Scenario Scope Value, (2) Posterior Action Value, and (3) Time Decay Value. Based on this, we introduce VAPS, a value-aware personalized search model that selectively incorporates high-value consultations through a consultation-user action interaction module and an explicit objective that aligns consultations with user actions. Experiments on both public and commercial datasets show that VAPS consistently outperforms baselines in both retrieval and ranking tasks.

Problem

• Assessing consultation value beyond semantic similarity for personalization
• Aligning consultations with user actions in personalized search
• Improving e-commerce search via value-aware consultation assessment

Innovation

• Framework assesses consultation value from three perspectives
• VAPS model integrates high-value consultations selectively
• Alignment of consultations with user actions improves search