Intelligent Approach for Attracting Churning Customers in Banking Industry Based on Collaborative Filtering

Elham Shafiei Gol, Abbas Ahmadi, Azadeh Mohebi

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review


During the last years, increased competition among banks has caused many developments in banking experiences and technology, while leading to even more churning customers due to their desire of having the best services. Therefore, it is an extremely significant issue for the banks to identify churning customers and attract them to the banking system again. In order to tackle this issue, this paper proposes a novel personalized collaborating filtering recommendation approach joint with the user clustering technology. In the proposed approach, first the loyal customers are clustered by means of hybrid algorithm based on Particle Swarm Optimization (PSO) and K-means. The clusters of loyal customers are then used to identify the features of the churning customers. Finally, the list of appropriate banking services are recommended for the churning customers based on a collaborative filtering recommendation system. The recommendation system uses the information of loyal customers to offer appropriate services for the churning customers. The proposed intelligent approach was successfully applied to return the churning customers of an Iranian bank.
TidsskriftJournal of Industrial and Systems Engineering
Udgave nummer4
Sider (fra-til)9-25
Antal sider17
StatusUdgivet - nov. 2016
Udgivet eksterntJa


  • Customer churn
  • Data clustering
  • Recommender system
  • Collaborative filtering
  • Particle swarm optimization