Results 1 -
5 of
5
1 How to Improve Group Homogeneity in Online Social Networks
"... Abstract — The formation and evolution of interest groups in Online Social Networks is driven by both the users ’ preferences and the choices of the groups ’ administrators. In this context, the notion of homogeneity of a social group is crucial: it accounts for determining the mutual similarity amo ..."
Abstract
- Add to MetaCart
(Show Context)
Abstract — The formation and evolution of interest groups in Online Social Networks is driven by both the users ’ preferences and the choices of the groups ’ administrators. In this context, the notion of homogeneity of a social group is crucial: it accounts for determining the mutual similarity among the members of a group and it’s often regarded as fundamental to determine the satisfaction of group members. In this paper we propose a group homogeneity measure that takes into account behavioral information of users, and an algorithm to optimize such a measure in a social network scenario by matching users and groups profiles. We provide an advantageous formulation of such framework by means of a fully-distributed multi-agent system. Experiments on simulated social network data clearly highlight the performance improvement brought by our approach.
An Integer Programming Approach and Visual Analysis for Detecting Hierarchical Community Structures in Social Networks
, 2015
"... Detecting community structures in social networks is a very important task in social network analysis as these community structures explain relationships among individuals and can be used to predict social behavior. The relation-ship among subcommunities in each community can further be identified a ..."
Abstract
- Add to MetaCart
(Show Context)
Detecting community structures in social networks is a very important task in social network analysis as these community structures explain relationships among individuals and can be used to predict social behavior. The relation-ship among subcommunities in each community can further be identified as hierarchical community structures, in which each super node at each hier-archical level represents a nested structure of communities or nodes. Most previous studies attempting to detect hierarchical community structures fo-cused on new metaheuristics that are computationally efficient but do not guarantee the optimal community partition. As a result, this work applies a novel integer programming (IP) approach to detect hierarchical commu-nity structures in social networks. This approach has flexible community capacity limits, does not limit the community numbers at different levels, and maximizes a quality measure for hierarchical community partition. The proposed IP approach can use existing software solvers to detect hierarchical community structures without implementing an algorithm. Visual analysis of experimental results shows that the proposed model with different set-tings for level numbers can analyze reasonable and sophisticated hierarchical community structures, such that the relationships between communities at different levels can be elucidated clearly.
A new Pre-processing Strategy for Improving Community Detection Algorithms
"... In the study of complex networks, a network is said to have community structure if it divides naturally into groups of nodes with dense connections within groups and only sparser connections between them. Detecting communities from complex networks has attracted attention of researchers in a wide ra ..."
Abstract
- Add to MetaCart
(Show Context)
In the study of complex networks, a network is said to have community structure if it divides naturally into groups of nodes with dense connections within groups and only sparser connections between them. Detecting communities from complex networks has attracted attention of researchers in a wide range of research areas, from biology to sociology and computer science. In this paper, we introduce a new approach to make existing community detection algorithms execute with better results. Our method enhancing community detection algorithms by applying a pre-processing step that exploits betweenness for nodes and edges, to maps unweighted graph onto a weighted graph. It has been tested in conjunction with four algorithms, namely the Louvain method, SOM algorithm, VOS clustering, and Danon algorithm. Experimental results show that our edge weighting strategies raises modularity for existing algorithms.
Mixing local and global information for community detection in large networks
"... The problem of clustering large complex networks plays a key role in several scientific fields ranging from Biology to Sociology and Computer Science. Many approaches to clustering complex networks are based on the idea of max-imizing a network modularity function. Some of these approaches can be cl ..."
Abstract
- Add to MetaCart
(Show Context)
The problem of clustering large complex networks plays a key role in several scientific fields ranging from Biology to Sociology and Computer Science. Many approaches to clustering complex networks are based on the idea of max-imizing a network modularity function. Some of these approaches can be classified as global because they exploit knowledge about the whole network topology to find clusters. Other approaches, instead, can be interpreted as local because they require only a partial knowledge of the network topology, e.g., the neighbors of a vertex. Global ap-proaches are able to achieve high values of modularity but they do not scale well on large networks and, therefore, they cannot be applied to analyze on-line social networks like Facebook or YouTube. In contrast, local approaches are fast and scale up to large, real-life networks, at the cost of poorer results than those achieved by local methods. In this article we propose a glocal method to maximizing modularity, i.e., our method uses information at the global level, yet its scalability on large networks is comparable to that of local methods. The proposed method is called COmplex Network CLUster DEtection (or, shortly, CONCLUDE.) It works in two stages: in the first stage it uses an information-propagation model, based on random and non-backtracking walks of finite length, to compute the importance of each edge in keeping the network connected (called edge centrality.) Then, edge centrality is used to map network vertices onto points of an Euclidean space and to compute distances between
Visualizing criminal networks reconstructed from mobile phone records
"... In the fight against the racketeering and terrorism, knowl-edge about the structure and the organization of criminal networks is of fundamental importance for both the investi-gations and the development of efficient strategies to prevent and restrain crimes. Intelligence agencies exploit informa-ti ..."
Abstract
- Add to MetaCart
(Show Context)
In the fight against the racketeering and terrorism, knowl-edge about the structure and the organization of criminal networks is of fundamental importance for both the investi-gations and the development of efficient strategies to prevent and restrain crimes. Intelligence agencies exploit informa-tion obtained from the analysis of large amounts of heteroge-neous data deriving from various informative sources includ-ing the records of phone traffic, the social networks, surveil-lance data, interview data, experiential police data, and po-lice intelligence files, to acquire knowledge about criminal networks and initiate accurate and destabilizing actions. In this context, visual representation techniques coordinate the exploration of the structure of the network together with the metrics of social network analysis. Nevertheless, the utility of visualization tools may become limited when the dimen-sion and the complexity of the system under analysis grow beyond certain terms. In this paper we show how we employ some interactive visualization techniques to represent crim-inal and terrorist networks reconstructed from phone traffic data, namely foci, fisheye and geo-mapping network layouts. These methods allow the exploration of the network through animated transitions among visualization models and local enlargement techniques in order to improve the comprehen-sion of interesting areas. By combining the features of the various visualization models it is possible to gain substantial enhancements with respect to classic visualization models, often unreadable in those cases of great complexity of the network.
