Company Networks Analysis. Meso-level structural analysis

      Аннотация: Our work deals with analysis of specific type of social networks called company networks. Actors are companies and connections are representing interactions between them. Dynamics of company networks is slower and they do not tend to create connections above certain boundaries (local business) because higher trust is needed than in other social networks. They also contain a lot of non-structural data as compositional data and temporal data. Company networks as interested already several projects in Slovakia like foaf.sk and ITLIS.eu projects. These projects focused on visualisation of ties in the network of companies in Slovak republic. These portals also partially utilize temporal data. In our work we build on the ITLIS.eu project and enhance it in a number of ways. We proposed a comprehensive data model of company networks, taking into account also aging factor (based on detailed temporal data). Main part of our work is dedicated to meso-level structural analysis, mainly analysis of local structures and local structure patterns detection, where approaches to large-scale network decomposition and positional analysis of local structures are designed and implemented.

Ключевые слова: Data Mining, social network analysis, local structures analysis, company network model