Improving the Compactness in Social Network Thematic Groups by Exploiting a Multi-dimensional User-to-Group Matching Algorithm

The internal organization of an Online Social Network is well described by the formation of groups between members. Often groups evolve in a very confusing way, due to occasional interactions between their components. However, it does not necessarily imply the formation of aggregation units in which users have similar interests and behaviour and, contextually, mutually trust with each others. Users can form groups (or join already existing groups) on the basis of shared interests or because dense social connections exist among group members, however, it is not uncommon the birth and growth of thematic groups, i.e., those groups arising from the social aggregation of users around a specific topics of interest. In this context, we argue that a multi-dimensional organization of the social network, in which each dimension represents the projection of the network on a given topic, could facilitate the task of forming compact groups. In this paper, after defining a notion of compactness for a group, that integrates similarity and mutual trust, we propose to provide each user with a software agent associated with each topic of interest for the user, and that represents a user's avatar in the corresponding dimension. This allows the user to delegate to his/her agent the management of group joining requests regarding a given topic, selecting only those interlocutors which appear the most appropriate for their owners. In our approach a Users-to-Group matching algorithm allows the agents to dynamically manage the evolution of the social network organization. Some experiments on real data clearly show the advantages introduced from our approach in assigning the users only to groups compatible with their orientations.