Rethinking Semantic Collaborative Integration: Why Alignment Is Not Enough

2026-04-23

AI Summary

This work addresses the limitations of existing approaches that integrate large language models (LLMs) with collaborative recommendation systems by overly relying on representation alignment while neglecting the heterogeneity and complementarity between semantic and collaborative views. To overcome this, the paper proposes a complementary fusion paradigm that treats the two modalities as heterogeneous views comprising both shared and private factors. The authors introduce a shared-private latent structure assumption and develop a suite of complementarity-aware diagnostic tools—including metrics such as overlap degree, unique hit contribution, and theoretical fusion upper bound—alongside controlled alignment probes for empirical analysis. Experiments on sparse recommendation benchmarks reveal low item-level consistency between the two views yet demonstrate substantial Oracle fusion gains, thereby validating their strong complementarity.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision

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
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Large language models (LLMs) have become an important semantic infrastructure for modern recommender systems. A prevailing paradigm integrates LLM-derived semantic embeddings with collaborative representations via representation alignment, implicitly assuming that the two views encode a shared latent entity and that stronger alignment yields better results. We formalize this assumption as the global low-complexity alignment hypothesis and argue that it is stronger than necessary and often structurally mismatched with real-world recommendation settings. We propose a complementary perspective in which semantic and collaborative representations are treated as partially shared yet fundamentally heterogeneous views, each containing both shared and view-specific factors. Under this shared-plus-private latent structure, enforcing global geometric alignment may distort local structure, suppress view-specific signals, and reduce informational diversity. To support this perspective, we develop complementarity-aware diagnostics that quantify overlap, unique-hit contribution, and theoretical fusion upper bounds. Empirical analyses on sparse recommendation benchmarks reveal low item-level agreement between semantic and collaborative views and substantial oracle fusion gains, indicating strong complementarity. Furthermore, controlled alignment probes show that low-capacity mappings capture only shared components and fail to recover full collaborative geometry, especially under distribution shift. These findings suggest that alignment should not be treated as the default integration principle. We advocate a shift from alignment-centric modeling to complementarity fusion-centric, complementarity-aware design, where shared factors are selectively integrated while private signals are preserved. This reframing provides a principled foundation for the next generation of LLM-enhanced recommender systems.

Problem

• semantic-collaborative integration
• representation alignment
• complementarity
• heterogeneous views
• recommender systems

Innovation

• complementarity-aware fusion
• semantic-collaborative integration
• shared-plus-private representation
• representation alignment critique
• LLM-enhanced recommendation