摘要
在分析用可拓思想研究非个性化内容推荐算法的可行性基础上,将可拓学中基元的思想引入到推荐系统中,实现了推荐系统定性和定量的结合.借助可拓学中"距"的思想,通过引入可推荐可拓集实现了内容推荐算法中相似性的计算,进而得到了一种新的内容推荐算法.最后通过数据计算验证了方法的可靠性和有效性.可拓思想的引入不仅为推荐系统的研究提供了新的思路,也为可拓学的研究拓宽了应用领域.
Based on the feasibility analysis of the extension thoughts used in the non- personalized content-based recommendation, basic element of the extenics is used in the recommendation system to get the aim of combining qualitative and quantitative. With the idea of ＂distance＂, the definition of the recommended extension set are used to compute the similarity between product and the recommended extension set, and then get a new algorithm for content-based recommendation. Finally, the effectiveness and validity of the methods are proved with examples. The research results not only provide a new idea to study the recommender system, but also enrich the application field of extenics.
出处
《系统科学与数学》
CSCD
北大核心
2015年第10期1209-1218,共10页
Journal of Systems Science and Mathematical Sciences
基金
河南省高等学校重点科研项目资助计划(15A120004)
河南科技厅基础研究项目(52300410208)
河南教育厅人文社科项目(2015-GH-056)资助课题
关键词
内容推荐算法
非个性化推荐
可拓学
基元
Content-based recommendation, non-personalized recommender systems,extensics, basic element.
