SAR: Semantic Analysis for Recommendation

TL;DR

SAR introduces a hierarchical semantic analysis method for recommendation systems, enabling semantic matching based on learned representations from user ratings, significantly improving recommendation quality.

Contribution

This paper presents a novel hierarchical semantic analysis approach that learns semantic representations from ratings, enhancing recommendation interpretability and accuracy.

Findings

SAR outperforms state-of-the-art baselines substantially

Semantic representations enable better matching between users and items

The method provides semantic interpretability in recommendations

Abstract

Recommendation system is a common demand in daily life and matrix completion is a widely adopted technique for this task. However, most matrix completion methods lack semantic interpretation and usually result in weak-semantic recommendations. To this end, this paper proposes a $S$emantic $A$nalysis approach for $R$ecommendation systems $(SAR)$, which applies a two-level hierarchical generative process that assigns semantic properties and categories for user and item. $SAR$ learns semantic representations of users/items merely from user ratings on items, which offers a new path to recommendation by semantic matching with the learned representations. Extensive experiments demonstrate $SAR$ outperforms other state-of-the-art baselines substantially.