计算机技术与发展2017，Vol.27Issue(1)：16-19,4.DOI:10.3969/j.issn.1673-629X.2017.01.004
基于用户偏好和项目特征的协同过滤推荐算法
A Collaborative Filtering Algorithm Based on Interest of User and Attributes of Item
摘要
Abstract
Using the analysis of project properties and user behavior provides the user with a valid solution recommended resources. It is recommended by the interest of the user and item’ s attributes. To the user,according to analysis of the user registration display attributes and the user’s behavior data (for browsing,viewing,downloading and sharing of project resources),as well as the quantization of the history of user behavior,the users could be divided into different neighbors. For the project,the clustering of project could divided into different resource types by its attributes and the evaluation of user to project. Therefore,the collaborative filtering algorithm is improved to solve the problems of single recommended results,little evaluation matrix data,low accuracy of recommendation as well as cold start for new users and new project. Recommended resources is achieved to change dynamically along with behavior and interest of users,to meet their requirements,achieving the purpose of the personalized recommendation,avoiding the waste of time for user to search resources in huge amounts of resources.关键词
协同过滤/推荐系统/用户属性/项目属性Key words
collaborative filtering/recommendation system/user attribute/item attribute分类
信息技术与安全科学引用本文复制引用
张应辉,司彩霞..基于用户偏好和项目特征的协同过滤推荐算法[J].计算机技术与发展,2017,27(1):16-19,4.基金项目
国家自然科学基金资助项目(61262058) （61262058）
