Empirical Study On The Spatial Characteristics Of Social Relationships And Interactions In Cyberspace

In recent years, social medium and social network services (SNSs) have become more and more popular around the world. The rapid development of SNSs has profoundly influenced many aspects of human society. During the passed decade, the society organizations and human activities have been fundermentally changed. The opportunity for potential human activity has gone beyond physical space to virtual space. Study on the formation mechanism of human relationships and the spatial-temporal characteristics of human behaviors in virtual space have been research hot spots in a wide variety of subjects such as cyber-sociology, behavior geography, human dynamics, and etc. However, due to metadata bottleneck, the research on human relationships or interactions and the research on the human spatial behaivors have been conducted separately for decades of years. There is still lack of the research on the spatial-temporal characteristics of the formation and evolution of social ties especially in cybersociety. Whether or not the physical factors still play great impacts on the human relations and interactions in virtual space as in physical space needs to be figured out.The main dissertation consists of four parts as follows.1. The study obtained, simulated, and then adjusted the probability distributions of the probability of human relationships and the probability density distributions of human interactions with distance in three types of environments based on the mass of data from a large social networking site, Tencent QQ. By using the single-variable factor analysis, the spatial factors have been verified that they are still alive and play great impacts on the human relationships and interactions in virtual space. However, the variation trends on the probability of intra-city and inter-city human relationships or interactions with physical distances are fundamentally different which could be called as "the spatial pattern of duality". In addition, distance-decay effects on the probability of human relationships and the probability density of human interactions both exist in virtual space; however, the probability density map of human interactions presents a concentric circles structure while the spatial probability distribution of the human relationships is more respected to the administrative borders.2. This study analyzed the spatial scale of the spatial distribution of the huma relationships based on the entropy theory of information. By measuring the entropy of each individual’s spatial distribution of friends in time-varying and space-varying conditions, this study found the potential spatial predictability in user making friends across the whole user base despite of the significant differences in online socialising.3. Based on the aggregated inter-person social relationships collected from the largest social network website in China------Sina micro-blog, an inter-city geographic social network (GSN) is built in this research. This study attempts to analyze the topological characteristics and spatial interaction model of the inter-city social relations in cybersocety of China. Results show that strong system and local heterogeneities exist in the inter-city GSN, social connectivity is significantly imbalanced among cities and city linkages in China. Distance factor works on the intensity of inter-city relations in cyber space, however, compared to the physical network, it was becoming weaker and weaker.4. This study made a comparision between the strong ties dominated social network (SSN) and the week ties dominated social network (WSN) from the perspective of the topology and geography at the microscale and macroscale by using the mass data from Tencent QQ and Sina Micro-blog website. The comparison results showed that from SSN to WSN, individuals are becoming much freer to form relationships and broad their circles, however the aggregated inter-city network is becoming more heterogeneous and imbalaced. Moreover, physical distance is becoming much weaker from WSN to SSN along with the deepening of globalization and the prosperous development of social media and SNSs.