基于改进VIKOR的大数据联盟数据资源群推荐方法研究

Group Recommendation Method of Big Data Alliance Data Resource Based on Improved VIKOR

【目的/意义】针对数据稀疏型用户推荐准确度低,大数据联盟群用户对群推荐结果整体满意度不高的问题,本文提出一种基于改进VIKOR的大数据联盟数据资源群推荐方法。【方法/过程】根据大数据联盟数据资源群用户特点,在构建群推荐矩阵时,将用户分为群内用户和群外用户,分别考虑不同用户评分对群推荐结果的影响;依据大数据联盟数据资源的特殊性,提出一种数据资源属性权重确定方法,对不同数据资源的各属性分别确权,从而提高群推荐质量。【结果/结论】实验结果表明,本文提出的算法不但能够为数据稀疏型用户提供较准确的推荐结果,而且有效提升了大数据联盟数据资源群用户的整体满意度。【创新/局限】本文将VIKOR算法改进后用于大数据联盟数据资源群推荐,有效改善了群推荐效果,但未考虑用户分群对群推荐结果的影响,接下来将对联盟用户如何准确分群进行研究。

【Purpose/significance】To solve the problems of low recommendation accuracy of data sparse users and low overall satisfaction of big data alliance group users with the group recommendation results,this paper proposes a recommendation method of big data alliance data resource group based on improved VIKOR.【Method/process】According to the characteristics of data resources group users of big data alliance,users are divided into in-group users and out-group users when constructing group recommendation matrix,and the influence of different user ratings on group recommendation results is considered respectively.According to the particularity of data resources of big data alliance,a method to determine the attribute weight of data resources is proposed to confirm the attributes of different data resources,so as to improve the quality of group recommendation.【Result/conclusion】Experimental results show that the algorithm proposed in this paper can not only provide accurate recommendation results for data sparse users,but also effectively improve the overall satisfaction of data resource group users of big data alliance.【Innovation/limitation】In this paper,the improved VIKOR algorithm is used in the group recommendation of data resource of the big data alliance,which effectively improves the effect of group recommendation,but does not consider the impact of user clustering on the results of group recommendation.Next,we will study how alliance users are clustered.