综合多维度优化的网络党建用户关系评价研究

User relations evaluation research on network party construction with integrated multidimensional optimization

针对云南省基层党建综合服务平台经典用户相似度算法结果精确度低的问题,提出一种党员用户关系评价模型.首先基于用户间的微博文本、位置、共同好友、交互、背景设计适用于该平台的相似度优化算法,然后利用最大似然估计方法综合5个维度的相似度结果,得到最终的党员用户关系评价模型.以平台真实数据对模型进行性能分析,结果表明,与基于网络距离和内容的相似度算法、基于微博的相似度算法相比,提出的优化算法及最终模型在准确率、召回率和F1值上均有较大提升.

A user connection evaluation model is presented in this paper for party members to improve the accuracy of the classic evaluation algorithm running on the integrated service platform for the grass-roots party construction of yunnan province.Firstly,an optimization algorithm is designed for this platform,in which the user’s microblog text,location,friends,interactions and background are taken into account.Then the overall similarity of the similarities of the above five dimensions is calculated by using the maximum likelihood estimation,and the proposed evaluation model is obtained.Finally,the real data on the platform is used to verify the performances.The results show that the accuracy,recall and F1 value of the suggested model are higher than those of network distance and content,the similarity algorithm based on Microblog.