Bi-NAS: Towards Effective and Personalized Explanation for Recommender Systems via Bi-Level Neural Architecture Search

2026-07-01

AI Summary

This work addresses the limitations of existing recommender systems, which often lack effective and personalized explanation mechanisms across diverse scenarios, thereby hindering user trust and decision quality. To overcome this, the authors propose Bi-NAS, a novel framework that introduces bilevel neural architecture search to recommendation explanation generation for the first time. Bi-NAS jointly optimizes cross-attention mechanisms and feature interaction functions, while leveraging the zero-shot prompting capabilities of large language models to automatically produce explanations aligned with both user intent and item attributes. Without requiring fine-tuning, the method enables co-optimization of model architecture and explanation quality, significantly improving recommendation accuracy and explanation effectiveness across four real-world datasets, thus delivering clear, reliable, and highly personalized justifications for recommendations.

Abstract

Recommender systems are vital in helping users navigate vast amounts of information, offering personalized suggestions and effective explanations for these recommendations. While previous efforts have attempted to provide such explanations, evaluating their effectiveness across various scenarios remains a challenge. Enhancing these explanations is essential for improving user engagement, trust, and decision-making. To facilitate effective explanations within the recommender system, we propose a Bi-level Neural Architecture Search (Bi-NAS) framework to optimize explanations. This approach simultaneously refines cross-attention mechanisms and feature interaction functions by exploring both intra-layer and inter-layer design spaces. Furthermore, we integrate Large Language Models (LLMs) to enhance explanation generation, leveraging zero-shot prompting to produce more effective and personalized justifications. By aligning user feature preferences with item quality scores, our approach ensures that explanations reflect both user intent and item attributes, improving transparency and reasoning depth. Extensive evaluations on four real-world datasets demonstrate that Bi-NAS not only boosts recommendation accuracy but also significantly improves the effectiveness of explanations for recommender systems, providing users with clear and reliable insights into the suggestions they receive.

Problem

• Recommender Systems
• Explainability
• Personalized Explanation
• Explanation Effectiveness
• User Trust

Innovation

• Bi-level Neural Architecture Search
• Explainable Recommender Systems
• Large Language Models
• Zero-shot Prompting
• Cross-attention Optimization