Geo-social Networking Data Visualization Analysis System:Design, Development And Application

With the popularity of location-enabled mobile intelligent devices and the rise of mobile Internet applications, more and more geo-social networking data have accumulated and illustrated great research value. There are many difficulties exist in current geo-social networking data analysis area, including huge data volume, complex data structure and difficult to display. Orienting to knowledge discovery from geo-social networking data and utilizing with information visualization techniques, this paper designs and develops a geo-social networks visualization analysis system, which contributes to industrial users related to geo-social networks and the public.The research contents and contributions of this thesis are presented as follows:1. Aiming with four aspects of geo-social networks, spatial data storage, information visualization and the related development techniques and tools, this paper studies and analyzes the theory and technology involved to build the system. First of all, the theory of geo-social network is studied. Then, the paper analyzes the geographical spatial data storage in two ways-relational databases and NoSQL databases and focuses on the NoSQL database MongoDB from the perspective of features, functions, vector data storage and usage scenario. Next, this paper describes principles of visualization methods in detail, including markercluster, HeatMap, parallel coordinates plot and node-link plot. Finally, techniques and tools to build the system are briefly expounded.2. Inspired by knowledge discovery from geo-social networking data, the thesis carries out requirements analysis, overall architecture design, data model design, core function modules design and visual analysis system development. According to the basic goal of system development, this paper analyzes the needs of the system from the perspective of user permissions, data storage, data query and the selection of visualization methods. Based on this research, the logical architecture and the physical structure of the system are proposed in this paper. The logical architecture of the system consists of the user layer, business logic layer and data storage layer. Then, the paper respectively designs corresponding data model for the basic information of system user, geographical location data, user relationship and other category data. From the aspects of data management, data retrieval, visualization and statistical analysis, this thesis designs functional modules of the system. Finally, the visualization analysis system of geo-social networking data realizes.3. A Case Study of geo-social networking data visual analysis on Foursquare is presented to demonstrate the availability and efficiency of the system. Firstly, this paper introduces the application background and gives a preliminary analysis of the data set. Then describes the cloud platform of GeoKSCloud. The contents include architecture and main function. On this basis, the system integrates and deploys to the GeoKSCloud. By means of dynamically linked views, this system combined multiple visualization analysis modules such as markercluster, HeatMap, node-link plot, parallel coordinates plot, data table and histograms, and provided comprehensive and effective analytical methods for the time, spatial, category distribution rule. Through the application examples, five laws which are of practical value are summarized. Application examples show that the visualization analysis system which is designed and realized by this paper have good availability and efficiency.