The Research On Social-Networking Users’ Behaviour Characteristics And Interior Mechanism

With the success of websites like Facebook and Renren.com etc., Social networking sites become well known as an integral part of internet life. In the high-speed development of social networking sites, users are the most precious resources, but the sites are facing serious risk of losing them in actual operations. Therefore, the social users’ behavior needs to be studied thoroughly. However, current academic research and practice mainly focused on users’ participation motives, and few did system research on the time statistics characteristics of users’ behavior.Based on this, this study commenced in accordance with the following lines:1、Based on behavior record data of Renren.com’s individual user homepage, empirical research was taken to study the time statistical properties of individual behavior and group interaction.The study found that the interevent time of users’ individual behavior and group interaction met the power-law distribution. Also, the power exponent of individual behavior’s interevent time distribution correlates positively with activity. The study proposes that the exponent of group interaction’s interevent time distribution correlates not only with activity, but also positively with social driving factor.2、According to the interpersonal activities of social networking features, this study established an interest-driven model which was influenced by social driving factor. According to the interpersonal activities of social networking features, this study established an interest-driven model which was influenced by social driving factor based on the interest-driven model established by SHANG Ming-Sheng, DAI Shuang-Xing et al.3、The model was effectively verified with MATLAB tools using actual data, and based on the empirical conclusions, advices were given on the product improvement and profit model of social networking site.The numerical simulation results consisted with the empirical results and the analysis results of model established. This study will not only effectively help social networking sites enhance the users’ viscosity and realize the profit pattern from multiple channels, but also be efficiently applied in practical work from another perspective.