一种基于AHP的智能电影推荐方法  

Intelligent movie recommendation method based on AHP

作　　者：郭宇红[1] 杨金然 

机构地区：[1]国际关系学院信息科技学院,北京100091

出　　处：《南京邮电大学学报（自然科学版）》2015年第3期97-105,112,共10页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：北京市人才培养共建项目;国际关系学院高等学校本科教学工程资助项目

摘　　要：电影网站提供了十分丰富的电影资源,然而用户必须花费大量时间搜索自己感兴趣的资源,如何帮助用户快速找到其想要的资源成为大型电影网站的重要需求。智能推荐技术的提出使满足这个需求成为可能。文中提出一种基于AHP(Analytic Hierarchy Process)层次分析法的智能电影推荐方法,综合考虑电影印象、故事、表演、导演、画面、音乐和类型等因素,通过用户对各因素两两比较、权重计算,生成个性化推荐列表。基于时光网的多个用户实证研究表明,结合AHP的智能电影推荐方法简单、有效,大部分推荐结果满足用户喜好。Film websites can provide abundant movie resources. However, users must spend a lot of time to search for contens they are interested in. How to find what users like is an important requirement in large film website searching. Intelligent recommendation technologies make satisfying the requirement possible. This paper proposes an intelligent movie recommendation method based on analytic hierarchy process （AHP）. By pairwise comparison of factors, such as the impression, plot, performance, director, frame, music and type of movie, and weight computation, the personalized recommendation list is produced. Multiple case studies of users based on website of mtime show that the movie recommendation method with AHP is simple, effective,thus most recommendation results can satisfy users＇ tastes.

关 键 词：层次分析法 电影推荐 电影排名 权重量化 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]