推荐系统中的隐语义模型研究

摘要

Latent factor model is an efifcient method for latent semantic analysis. The core idea of latent factor model is to link users and items through implicit features. In this paper, the theory of latent factor model was introduced ifrstly, and its representation and solution was discussed subsequently. Three classic Latent factor models were introduced in details. Furthermore, the performances of these models were compared through experiments based on test dataset, and the characteristic of latent factor model in applications were also discussed.%隐语义模型是一种有效的隐含语义分析技术，其核心思想是通过潜在特征联系用户和物品。本文从理论方法的角度，详细介绍了隐语义模型的工作原理、模型表示、学习方法和评价指标等，并通过具体实验分别讨论几种典型隐语义模型算法的推荐效果，以全面认识和理解该模型在推荐系统中的应用特点。