基于加权关联规则的个性化推荐研究

STUDY ON PERSONALIZED RECOMMENDATION BASED ON WEIGHTED ASSOCIATION RULE

关联规则是个性化推荐系统中最重要的技术手段之一。传统的基于关联规则的个性化推荐认为,每个项目都具有相同的重要性,在实际应用中缺乏一定的针对性。在New-Apriori算法的加权支持度基础上结合Fp-growth算法思想,提出了基于Fp-树的加权关联规则算法。在实验中采用网页被用户选择的频率作为权重值,在个性化推荐系统中对该算法进行了实现。实验结果表明该算法具有较高的准确性和效率。

Association rule is one of the most important techniques in personalized recommendation system. Conventional personalized recommendation based on association rule considers that every item is of the same importantance,and the recommendation lacks some pertinency. Based on the weighted support of New-Apriori algorithm and with the combination of Fp-growth algorithm, a weighted association rule algorithm based on Fp-tree is put forward. In experiment ,the visit frequency of web page is used as weight ,and the algorithm is implemented in the personalized recommendation system. The experimental results show that the algorithm has high accuracy and efficiency.