Skyline Community Search in Multi-valued Networks

Abstract

Given a scientific collaboration network, how can we find a group of collaborators with high research indicator (e.g., h-index) and diverse research interests? Given a social network, how can we identify the communities that have high influence (e.g., PageRank) and also have similar interests to a specified user? In such settings, the network can be modeled as a multi-valued network where each node has d (d ≥ 1) numerical attributes (i.e., h-index, diversity, PageRank, similarity score, etc.). In the multi-valued network, we want to find communities that are not dominated by the other communities in terms of d numerical attributes. Most existing community search algorithms either completely ignore the numerical attributes or only consider one numerical attribute of the nodes. To capture d numerical attributes, we propose a novel community model, called skyline community, based on the concepts of k-core and skyline. A skyline community is a maximal connected k-core that cannot be dominated by the other connected k-cores in the d-dimensional attribute space. We develop an elegant space-partition algorithm to efficiently compute the skyline communities. Two striking advantages of our algorithm are that (1) its time complexity relies mainly on the size of the answer s (i.e., the number of skyline communities), thus it is very efficient if s is small; and (2) it can progressively output the skyline communities, which is very useful for applications that only require part of the skyline communities. Extensive experiments on both synthetic and real-world networks demonstrate the efficiency, scalability, and effectiveness of the proposed algorithm.

A survey of community search over big graphs

“…Simple graphs Attributed graphs Keyword Location Temporal Influence (weight) Profile k-core [175,46,15,66] (P. 1, 2, 3, 4, 5) [61,58] (P. 6) [60,65,185,221] (P. 7,8,9) [129] (P. 10) [127,128,30,215,21,126] (P. 12,13) [31] (P. 14) k-truss [98,6,101] (P. 15,16) [102] (P. 17) -- [216] (P. 18) k-clique [45,205,195,187] (P. 19,20,21,22) -- [125] Example 2 Let us reconsider the graph G in Fig. 2(a).…”

“…The model studied in [126], called the skyline community search, is based on the one-dimensional influential community model proposed in [127]. The authors defined the value of H on the Fig.…”

“…In [126], the authors first developed an efficient algorithm, called SkylineComm2D, to find all the skyline communities in the 2D case, i.e., d = 2. The time complexity of SkylineComm2D is O(s(m + n)) where s denotes the number of 2D skyline communities (i.e., the cost, the space cost of keeping such an index is O( l · n), where l denotes the average size of W (v) over V .…”

The core decomposition of networks: theory, algorithms and applications

“…In [91] the concept of k-core is combined with the concept of skyline, in order to spot communities that are not dominated. This technique is applied to graphs with d numeric attributes per node.…”

Community Focusing: Yet Another Query-Dependent Community Detection

“…Given a set of query nodes, studies (Sozio and Gionis 2010; Barbieri et al 2015;Huang et al 2015;Wu et al 2015;Yuan et al 2018) discover a densely connected subgraph containing the nodes. Some variants of community search, such as attributed (Huang and Lakshmanan 2017;Fang et al 2016;Shang et al 2017), influential (Li et al 2015), spatial (Fang et al 2017) and skyline (Li et al 2018) community search, have also been studied recently.…”