现代电子技术2017，Vol.40Issue(5)：78-81,4.DOI:10.16652/j.issn.1004-373x.2017.05.020
基于个性化特征的协同过滤推荐算法
Collaborative filtering recommendation algorithm based on personalized feature
摘要
Abstract
Since the recommendation method of the similarity feature point has low matching degree for the user′s persona-lized demand,a collaborative filtering recommendation algorithm based on personalized feature is put forward. The scoring model and project attribute model of the user information were constructed by taking the social network as the structure model. The trust degree conditional probability analysis method is adopted to construct the reliability recommendation mode. And then the personalized features are analyzed and extracted to match the personalized feature demand and project interest point reasonably, so as to implement the collaborative filtering recommendation. The test analysis was conducted with simulation experiments. The test results show that the method has high user scoring to perform the collaborative filtering recommendation of the social net-work project,the mean absolute error and root mean square error are small,and the recommendation quality is improved.关键词
个性化特征/协同过滤推荐/评分模型/项目属性Key words
personalized feature/collaborative filtering recommendation/scoring model/project attribute分类
信息技术与安全科学引用本文复制引用
李涛..基于个性化特征的协同过滤推荐算法[J].现代电子技术,2017,40(5):78-81,4.基金项目
贵州省科技技术基金项目(黔科合LH字[2014]7439号) （黔科合LH字[2014]7439号）
