摘要
视频搜索是目前信息检索领域研究的热点.提出一种利用协同过滤技术来实现个性化的视频搜索.该算法根据用户项目兴趣相似度来计算目标项目的得分,从而为每个用户产生一个推荐列表.实验结果表明该排名算法较IMDB搜索和Google搜索的结果在用户满意度上有很明显的提高.
Video search is a hotspot in current information searching field. The paper presents a new video search by u- sing collaborative filtering technique to realize individuality. The algorithm computes the scores of the target items ac- cording to the similarity degree of the items in which the users are interested, and thus generates a recommendation list for every user. The experimental results show that compared with IMDB search and Google search, our algorithm can obtain better rank results and improve user's satisfaction highly.
出处
《南京师范大学学报(工程技术版)》
CAS
2008年第4期182-185,共4页
Journal of Nanjing Normal University(Engineering and Technology Edition)
关键词
协同过滤
推荐
视频搜索
个性化
collaborative filtering, recommender, Video search, individuation
作者简介
通讯联系人:李慧,硕士研究生,研究方向:数据挖掘、智能信息系统.E-mial:shufanzs@126.com