Title: A new ego network model and an approach to extracting an ego network compliant with this model from a social internetworking system
Authors: Antonino Nocera; Domenico Ursino
Addresses: DIMET, Università Mediterranea di Reggio Calabria, Via Graziella, Località Feo di Vito, 89122 Reggio Calabria, Italy ' DIMET, Università Mediterranea di Reggio Calabria, Via Graziella, Località Feo di Vito, 89122 Reggio Calabria, Italy
Abstract: In this paper, we propose a new ego network model and an approach to extracting an ego network compliant with this model from a social internetworking system. Ego networks have been largely exploited in the past in the context of social network analysis where they referred to single social networks. Actually, we argue that they can also be profitably exploited in many application fields different from social network analysis, such as resource and user recommendation, extraction of hidden relationships among users and resources, definition of more cohesive communities. Moreover, they could refer to a set of social networks interacting with each other (i.e., a social internetworking system), instead of to a single social network. In this paper, first we introduce a new model to represent a social internetworking system, and a new ego network model compliant with a social internetworking scenario and conceived to favour the exploitation of ego networks in several new application fields. Then, we describe our ego network extraction approach. After this, we illustrate some experiments devoted to test our approach. Finally, we examine some related approaches and highlight the similarities and the differences with our own.
Keywords: ego networks; social networks; recommender systems; user recommendations; recommendation systems; web based communities; virtual communities; online communities.
International Journal of Web Based Communities, 2013 Vol.9 No.4, pp.483 - 518
Available online: 13 Oct 2013 *Full-text access for editors Access for subscribers Purchase this article Comment on this article