基于页面聚类的推荐算法的改进

AMELIORATION OF RECOMMENDATION ALGORITHMIC BASED ON PAGE CLUSTERING

基于页面聚类的推荐算法常被应用在个性化推荐系统中,但是很少考虑页面访问的顺序性。针对这种弊端,提出了一种新的路径相似度系数,同时在推荐算法中运用了关联规则,提高了推荐结果的准确性。

Recommendation algorithmic based on page clustering is often used in personalized recommendation systems, yet accessing order isn＇t considered in this algorithmic. To address this limitation, a new similarity coefficient is proposed. At the same time, associate rule is used in the recommendation algorithmic, which improves the accuracy of the recommended results.