Developing a Mining Methodology for Customer Relationship Analytics

Abstract

Online social networks emergence enables individuals to communicate and share their opinions and feedback. Social networks mining aims at applying mining techniques to online social networks to reveal interesting hidden patterns on human behavior and interaction. Organizations utilize social network mining to expand their market and to improve customer social relations.

In this work, we propose a novel social network mining approach for social customer relationship analysis. In our approach, we propose a community detection technique which benefits from the most influential users on the network. As communities tend to be formed around users of great influence on their peers, the proposed approach utilizes such influential users to build communities around them.

Moreover, we propose a new community detection algorithm that incorporates behavioral information attached to users in the social network. Using such behavioral data of nodes is for the aim of detecting communities that are closely mapped to the underlying behavioral communities in real social networks. We use the behavioral data, namely, the actions done by users on their social network to propose a new similarity measure to measure the degree of similarity between users. Furthermore,the proposed algorithm uses the demographic data of users to enhance the quality of communities detected.

Experimental evaluation on two real social network datasets has been carried out and the results show that the proposed social network mining approach surpasses others in respect of all the evaluation measures used which indicates the ability ofthe proposed approach in identifying communities with high quality.