基于活动领域分类与间接关系融合的社会化网络用户关系强度计算模型

User Relationship Strength Estimation Model In Online Social Networks based on Fusion of Activity Field Classification and Indirect Relationship

在社会化网络中用户关系强度是网络营销的重要依据之一,用户关系强度的精确计算能够提高推荐的针对性与准确性,实现精准营销。本文提出一种基于活动领域分类与间接关系融合的社会化网络用户关系强度计算模型。该模型首先采用LDA算法与相似度计算的方法将用户交互活动文档分为工作、饮食、购物、旅游、运动、娱乐和其他7个活动领域。然后以交互活动文档为依据计算各个领域中用户的单向综合关系强度,其中包括直接关系与间接关系,用社交网络权重图表示。结合交互频率与时间的影响因素计算直接关系强度,结合关系路径长度、关系路径数量和关系路径的边权重计算间接关系强度。以新浪微博为研究对象,实验结果表明,该方法相比于一般的社会化网络用户关系强度计算方法更加精确。

The user relationship strength in online social network is one of the important basis of network marketing. The accurate calculation of it can improve the pertinence and the accuracy of recommendation, implementing precision marketing. This paper proposed a estimation model of user relationship strength in online social networks based on the fusion of activity field classification and indirect relationship method. At first, using LDA algorithm and similarity calculation method divide user interaction activity documents into 7 activity fields, including working, diet, shopping, traveling, sports, entertainment and others. Then According to the interaction activities document to calculate the user＇s one-way comprehensive strength in each field. Which including direct relationship and indirect relationship, and indicated by social network weights diagram. Combining the affecting of interactive frequency with time to calculate direct relationship strength, combining the length of a relationship path, the number of relationship paths with the edge weights along with a relationship path to calculate the indirect relationship strength. Selecting sina weibo as the research object, the experimental results show that the method is more accurate.