Inference Computation Scaling for Feature Augmentation in Recommendation Systems

2025-02-22

AI Summary

Existing recommender systems suffer from incomplete feature coverage and insufficient descriptive specificity in fast-inference-based feature enhancement, hindering fine-grained user preference modeling. This paper pioneers the integration of reasoning scaling into recommendation-oriented feature enhancement: by extending large language models’ Chain-of-Thought (CoT) reasoning paths, optimizing search strategies, and generating high-quality semantic features, it systematically improves both feature diversity and semantic precision. Experiments demonstrate a 12% improvement in NDCG@10, confirming a positive correlation between reasoning depth and recommendation performance. The core contribution lies in uncovering the dual-gain mechanism—where CoT path extension simultaneously amplifies feature quantity and specificity—thereby establishing a scalable, LLM-powered feature enhancement paradigm for recommender systems.

Technology Category

• Machine Learning: Feature Construction/Reformulation
• Data Mining & Knowledge Management: Recommender Systems
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Large language models have become a powerful method for feature augmentation in recommendation systems. However, existing approaches relying on quick inference often suffer from incomplete feature coverage and insufficient specificity in feature descriptions, limiting their ability to capture fine-grained user preferences and undermining overall performance. Motivated by the recent success of inference scaling in math and coding tasks, we explore whether scaling inference can address these limitations and enhance feature quality. Our experiments show that scaling inference leads to significant improvements in recommendation performance, with a 12% increase in NDCG@10. The gains can be attributed to two key factors: feature quantity and specificity. In particular, models using extended Chain-of-Thought (CoT) reasoning generate a greater number of detailed and precise features, offering deeper insights into user preferences and overcoming the limitations of quick inference. We further investigate the factors influencing feature quantity, revealing that model choice and search strategy play critical roles in generating a richer and more diverse feature set. This is the first work to apply inference scaling to feature augmentation in recommendation systems, bridging advances in reasoning tasks to enhance personalized recommendation.

Problem

• Scaling inference for feature augmentation
• Improving feature specificity in recommendations
• Enhancing user preference capture with CoT reasoning

Innovation

• Scaling inference improves feature quality
• Extended Chain-of-Thought enhances feature precision
• Model choice impacts feature diversity generation