Abstract: In the field of computer science, topics regarding analysis of social networks grows more and more popular. It stems from the fact that the knowledge hidden in connections between people might enable us to answer many questions concerning various areas of life. When analyzing social networks, we are usually interested in such information as – network structure, how it is managed and what is the scheme of flow of goods or information. When starting a study of processes taking place in social networks it can be noticed that a lot of data subject to analysis is stored in structures of databases. Thus, database server’s capabilities in terms of social network examining and assigning roles to its members are worth analysing. The aim of this paper is to present studies regarding using a database serve to gather data on social networks and mechanisms it provides to specify the type of connections between network elements. One of the tasks of social network analysis is a graphic visualization of connections between various objects existing in the field of interest of a conducted research. The most convenient representation of such is a graph. Ensuring a possibility to gather data for a permanent storage and at the same time effectively operate on them required designing appropriate data structures. It was decided, that the relational database meets the demands. The basic structure of the proposed data model covers vital graph elements along with a description of their qualities. Analysis of social networks presented as graphs poses a challenge when it consists of a huge number of nodes and connections between them. One of the most important functions for presenting a social network is to make it possible to locate and visualize only fragments of the whole graph in order to make analyses more effective. In the research, the mechanism for limiting set of nodes (creating sub-graphs, called paths for the purpose of this research) and for aggregating elements of a graph were prepared. Moreover, problem of graph search was also taken into consideration. The two algorithms, BFS and Dijkstra's, were implemented as a database procedures. Effectiveness of the operation of the first of algorithms was compared with its implementation in the. NET environment as well.

Keywords: social network analysis, database server, graphs, shortest paths.

ACM Classification Keywords: H. Information Systems H.3 INFORMATION STORAGE AND RETRIEVAL H.3.3 Information Search and Retrieval

Link:

DATABASE SERVER USAGE IN THE SOCIAL NETWORKS ANALYSIS

Katarzyna Harężlak

http://foibg.com/ibs_isc/ibs-23/ibs-23-p04.pdf