Fast and Improved Clustering Technique with User Profile Information for Correlated Probabilistic Graphs

Objectives: The main objective of this work is to achieve the better efficiency and accuracy of the clustering along with the user profile information.

Methods: Partially Expected Edit Distance Reduction (PEEDR) technique is used for adding or eliminating vertices from the clusters. Correlated Probabilistic Graph Spectral (CPGS) is used to progress the quality of cluster. Improved attractiveness-based community clustering is a weighted clustering approach which enhances the clustering performance in superior.

Findings: The proposed method achieves high performance in terms of precision, recall and accuracy.

Application/Improvements: The proposed system is done by using improved attractiveness-based community clustering (IACC). It performs the clustering process based on the weight value node and edge in the network. The weight of node implies the core degree of the person in the network, and the weight of edge means the attractiveness between the two nodes. Additionally, this method performs the efficient graph clustering technique which combines the user profile of users.