Efficient Detection of Overlapping Communities Using Asymmetric Triangle Cuts

Tags: , ,


The aim of the project is community does not overlapping the social network then communities intermediate will be revoke the community does not perform this problem also resolve this project.

Existing System:

Social network communities are overlapping because the user join us multiple community and posted file like whatapp, facebook its one posted share with multiple groups it’s the group member join so many group and posted group. The member watch the file multiple group and it’s also download the file so memory space increasing. Social networks can be clustered into cohesive groups called communities, where the vertices within a community are densely connected with each other, while they are sparsely connected to the vertices outside the community. For example, for targeted advertisements in a consumer network, detecting overlapping communities helps to identify groups of members with similar shopping preferences, and they will become suitable audiences for an advertisement campaign, as they usually share several shopping preferences.

Proposed System:

We here proposed a new fitness metric, overlapping triangle connectivity, for overlapping community detection that can minimize free rider and separation effects on overlapping communities We first introduce a new definition of internal density and external sparsely of communities based on ‘asymmetric triangle cuts’, and propose a new fitness metric for overlapping community detection that mitigates both free rider and separation effects on communities. Therefore, good community fitness metric should favor a large number of triangles within a community. Inversely, the connectivity between communities is not a preferred property, since a large number of edges between two communities may not represent strong connections of vertices between the two communities.