Classification and visualization of bibliography data in computer sciences

Fast and ever-growing communities form big data sources. To examine the structure of communities and interactions between them, it is necessary to analyze these data sources with various scientific methods in order to obtain healthy information about the communities and to observe their development and change. Social networks are one of the important structures used to model and visualize communities and their relationships. By modeling large and complex systems in the form of social networks, it is possible to make their structures more understandable and to analyze their connections and changes. In addition, methods of extracting information are used to make the structure of the communities more understandable and to improve the structure and dynamic of the network. By using information extraction methods, it is possible to obtain information and connections that are not yet known from different sources. Academic communities are examples of rapidly growing communities. Social networking structures and methods of information extraction are important methods used in the analysis of academic communities. In this thesis, three basic studies are presented. The first study is the simplification and integration of academic community data from different sources. For this purpose, DBLP, ACMCCS and ASN academic community data sources were used. Information extraction methods were applied to the data obtained and the missing links within the network were determined and the structure of the network to be formed was improved. The second study is the examination of the assets in the academic community and their connections. For this purpose, the community is modeled as a network. Subnets formed by authors, publications and other assets were identified. Classification is made according to the characteristics of the network. Finally, an online web application has been developed where the results obtained can be queried, monitored and reported.