情境化条件下基于OBN模型的推荐算法研究

OBN-based recommended algorithm research under contextualization

随着互联网的发展,用户获取信息精准度提高,个性化服务越来越重要.针对个性化推荐算法中用户相似性计算精度不高,导致算法的推荐精度低的问题,应用面向对象思想和贝叶斯理论、融合推荐对象的情境化因素,提出OBN模型,实现用户之间的相似度计算方法 ,该方法具有时间复杂度低、聚类稳定性强的特点.在此基础上设计个性化推荐算法.通过实验分析,提高个性化推荐的精度.

With the development of the Internet, users require information with higher accuracy, which makes personalized service more and more important. In order to improve the low accuracy of calculating the users＂ similarity by personalized recommended algorithm and the low recommendation accuracy problem, this paper puts forward the OBN method by applying the object-oriented thinking and Bayesian theory and combining the situational factors of the objects to figure out the algorithm that can calculate the similarities between users. The method has the characteristics of low time complexity and strong clustering stability, on which the personalized recommended algorithm is designed. And through experimental analysis, we can improve the personalized recommendation accuracy.