Abstract: It is stable for the development of content-based recommendation system to represent the features of merchandise.The theory of Vague sets，Gaussian function and characteristics of uncertainty are used to represent features with Vague value.On this basis，the general steps of content-based recommendation with Vague sets are given，in order to get a new idea and method to recommender systems designers.Finally，some recommender formulas with different features are given，which will be conducive to the work of the actual recommendation.Select different formulas according different conditions will improve the quality and accuracy of the recommendation.

Key words: recommender systems, Vague sets, similarity, content-based recommendation

摘要： 基于内容推荐系统发展的瓶颈在于商品特征的表示。借助Vague集方法，并利用高斯函数和Vague集中未知度的一些性质，将商品的特征用Vague值形象表示。在此基础上，给出了运用Vague集方法进行基于内容推荐的步骤，为推荐系统的设计提供了一种新的指导思想和方法。最后给出了多个推荐公式，并进行简单的比较，有利于在实际推荐工作中酌情选择不同的推荐公式，提高商品推荐质量和推荐精度。

关键词: 推荐系统, Vague集, 相似度, 基于内容推荐