| 112 | 0 | 59 |
| 下载次数 | 被引频次 | 阅读次数 |
随着电子商务产业的不断发展,推荐系统越来越多走入人们的生活,其中Top-K推荐能够推荐一个商品列表供用户选择,在商业推荐中越来越多地扮演重要角色。对于Top-K推荐而言,多样性的提高可以使推荐列表不再重复、单调,给用户多样化的选择空间,更容易适应用户需求。传统的Top-K推荐方法大多在预测评分方法的基础上进行优化改良,本文通过MovieLens数据集上的统计调查,说明基于用户兴趣分布会比基于预测评分拥有更优的效果。本文还提出两个获取用户兴趣分布的思路,供后续研究参考。
Abstract：With the continuous development of e-commerce industry,recommender systems go more and more into people's lives.Top-K recommendation could recommend a list of items for the user to choose,and more and more plays an important role in the commercial recommendation.For Top-K recommendation,the increase in diversity can make the recommendation lists no longer repeat and monotonous.It provides users a variety of choices,convinience to adapt to users' needs.The traditional Top-K recommendation methods are mostly optimized on the basis of the rate prediction methods.Through the statistical survey on the MovieLens dataset,this paper shows that the user interest distributions will have better effect than the predictive rate.This paper also proposes two ideas for obtaining user interest distributions which may be used in later research.
[1]Koren Y,Bell R.Advances in Collaborative Filtering.Recommender Systems Handbook[M].2011:145-186.
[2]Koren Y.The Bellkor Solution to the Netflix Grand Prize[J].Netflix prize documentation,2009(81):1-10.
[3]Mi Zhang,Neil Hurley.Avoiding Monotony:Improving the Diversity of Recommendation Lists[R].Lausanne,Switzerland,October,2008:23-25.
[4]Zhou T,Kuscsik Z,Liu J G,et al.Solving the Apparent Diversity-accuracy Dilemma of Recommender Systems[J].Proceedings of the National Academy of Sciences of the United States of America,2010,107:4511-4515.
[5]Qin Lijing,Zhu Xiaoyan.Promoting Diversity in Recommendation by Entropy Regularizer[J].IJCAI,2013,2698-2704.
[6]Jaime G.Carbonell,Jade Goldstein.The Use of MMR,Diversity-based Reranking for Reordering Documents and Producing Summaries[C].SIGIR1998,Melbourne,Australia,
[7]Tommaso Di Noia,Vito Claudio Ostuni,Jessica Rosati,et al.An Analysis of Users'Propensity Toward Diversity in Recommendations[C].Eighth ACM Conference on Recommender Systems,Silicon Valley,2014.
[8]Rodrygo L T.Santos,Craig Macdonald,IadhOunis.Exploiting Query Reformulations for Web Search Result Diversification[C].Raleigh,2010.
[9]A.Ashkan,B.Kveton,S.Berkovsky,et al.Optimal Greedy Diversity for Recommendation[C].Proceedings of the Twenty-Fourth International Joint Conference on Artificial Intelligence,2015,1742-1748.
[10]Van Dang,W.Bruce Croft.Diversity by Proportionality:An Election-based Approach to Search Result Diversification[C].SIGIR12,Portland,2012:12-16.
[11]Deepak Agarwal,Bee Chung Chen.FLDA:Matrix Factorization through Latent Dirichlet Allocation[C].WSDM 2010.
[12]Zhang Y,Wu J,Zhuang Y.Random Walk Models for Top-N Recommendation task[J].Journal of Zhejiang University(SCIENCE A),2009,10(7):927-936.
基本信息:
中图分类号:TP391.3
引用信息:
[1]邢小璐.Top-K推荐中的多样性研究[J].微型电脑应用,2017,33(09):44-46.
2017-09-20
2017-09-20