Remy Cazabet

Detection of Overlapping Communities in Dynamical Social Networks

Community detection on networks is a well-known problem encountered in many fields, for which the existing algorithms are inefficient 1) at capturing overlaps in-between communities, 2) at detecting communities having disparities in size and density 3) at taking into account the networks’ dynamics. In this paper, we propose a new algorithm (iLCD) for community detection using a radically new approach. Taking into account the dynamics of the network, it is designed for the detection of strongly overlapping communities. We first explain the main principles underlying the iLCD algorithm, introducing the two notions of intrinsic communities and longitudinal detection, and detail the algorithm. Then, we illustrate its efficiency in the case of a citation network, and then compare it with existing most efficient algorithms using a standard generator of community-based networks, the LFR benchmark.

Simulate to Detect: A Multi-agent System for Community Detection

Community detection in social networks is a well-known problem encountered in many fields. Many traditional algorithms have been proposed to solve it, with recurrent problems: impossibility to deal with dynamic networks, sensitivity to noise, no detection of overlapping communities, exponential running time. This paper proposes a multi-agent system that replays the evolution of a network and, in the same time, reproduces the rise and fall of communities. After presenting the strengths and weaknesses of existing community detection algorithms, we describe the multi-agent system we propose. Then, we compare our solution with existing works, and show some advantages of our method, in particular the possibility to dynamically detect the communities.

Automated community detection on social networks: useful? efficient? asking the users

In most online social networks, with the increasing number of users and content, the problem of contact filtering becomes more and more present. The recent introduction of such features in online social networks -- for instance, Circles in Google+ or Facebook Smart lists -- shows that it is a problem they are confronted to. In this paper, we explore this question through multidisciplinary aspects. First, we discuss about this issue of groups management in the context of social networks. Then, we present several techniques from the state of the art to automatically find meaningful groups of contacts in a users contact list. Finally, we asked Facebook users to evaluate these solutions on their own Facebook network, both to compare the solutions among themselves and to assess how pertinent the best ones are according to them. The conclusions of this study is that a network analysis approach can strongly improve the efficiency of an automated detection of groups on networks, which could be used, combined with profile data extraction, to design intelligent management of groups of contacts.

Characterizing the nature of interactions for cooperative creation in online social networks

Many aspects of online social networks (OSN) have been studied in recent years. In this article, we focus on the question of interactions in large OSN. We propose methods to study these interactions, and apply them on a platform called Nico Nico Douga (NND), with the aim of understanding cooperative behaviors, taking the form of collective creation of music videos in NND. Our first contribution is a method that, starting from the network of interactions between users, evaluates three aspects: the impact of the social structure on these interactions, their concentration, and their reciprocity. We characterize the nature of interactions in NND, and compare it with four different datasets. We find that interactions in NND are more similar to a diffusion process, such as retweets in Tweeter, than to interpersonal communications, or even to cooperation in science. Our second contribution is a typology of roles for productions in a cooperative process. These roles are attributed based on the neighborhood of the nodes in the network of references between productions. We define direct roles, relative roles, and indirect roles. We subsequently study the frequency of these roles in NND. We show a correlation between the category of the contribution of a video (song, animation, etc.) and its probability of having a certain role. We also find a positive correlation between the most active users and the production of videos playing an important role in the cooperation process.

Understanding massive artistic cooperation: the case of Nico Nico Douga

Many online social networks have been studied in the last decade, giving us insights into the way people diffuse information, communicate, and organize themselves. In this article, we focus on the emergent organization in massive artistic cooperation. We study the creation process of complex music videos in a platform called Nico Nico Douga. We give insights into three aspects of emergent organization:The relation between popularity (in terms of view) and influence on the cooperation process.The specialization of creators.The organization of the network of citation.