Start Date

16-8-2018 12:00 AM

Description

Multi-Relational networks are pervading in many fields such as technological products, twitter, and bibliographic websites. Many studies have focused on matters related to detecting virtual communities from social networks. Nevertheless, most of them have focused on analyzing single relational networks. Some communities discovering methods from multi-relational networks convert them to single-relational networks first and assume that different relations were independent from each other, which is obviously unreal in the real-life cases. In this paper, we attempt to address this challenge by introducing a new approach called CoMRCA for community detection from multi relational networks which incorporate the multiple types of objects and relationships. First, we have proposed to construct the Concept Lattice Family (CLF) aiming to model the different objects and relations of the multi-relational network using the Relational Concept Analysis (RCA). In the second step, we have proposed a new algorithm called SearchCommunity to navigate between the set of lattices and extract the multi-relational communities and their appropriate labels. Carried out experiments on real-datasets enhance the effectiveness of our proposal and lead to promising opportunities.

Share

COinS
 
Aug 16th, 12:00 AM

Relational concept analysis for virtual community detection from multi-relational networks

Multi-Relational networks are pervading in many fields such as technological products, twitter, and bibliographic websites. Many studies have focused on matters related to detecting virtual communities from social networks. Nevertheless, most of them have focused on analyzing single relational networks. Some communities discovering methods from multi-relational networks convert them to single-relational networks first and assume that different relations were independent from each other, which is obviously unreal in the real-life cases. In this paper, we attempt to address this challenge by introducing a new approach called CoMRCA for community detection from multi relational networks which incorporate the multiple types of objects and relationships. First, we have proposed to construct the Concept Lattice Family (CLF) aiming to model the different objects and relations of the multi-relational network using the Relational Concept Analysis (RCA). In the second step, we have proposed a new algorithm called SearchCommunity to navigate between the set of lattices and extract the multi-relational communities and their appropriate labels. Carried out experiments on real-datasets enhance the effectiveness of our proposal and lead to promising opportunities.