Mechanical Engineering & Center for Energy,
Environment and Economy
Customer Churn Prediction Using Social Network Analysis in a
Turkish Bank Setting
ABSTRACT
The main goal of this joint project is to exchange knowledge and expertise and to transfer technology among the involved
research institutions within an industrial application driven setting. More specifically, the project aims to develop and
implement relational learners to predict customer churn using social network information in a banking context.
Huge amounts of networked data on a broad range of network processes and information flows between interlinked entities
are available, such as for instance money transfers connecting bank accounts, or call logs linking telephone accounts.
These massive data logs potentially hide information that is extremely valuable to companies and organizations, but as well
is extremely difficult to discover due to the size and the fragmentation of the data. Building upon an established research
experience in the fields of data mining and relational learning, marketing modeling in the financial sector, and customer
churn prediction, the main objectives of this project are:
• RO1: to implement and develop scalable versions of current relational learning techniques that are applicable to bank
transfer data
• RO2: application of social network analysis to improve the performance of customer churn prediction models using stateof-the-art relational learners in a real-life banking setting
Doç. Dr.
Ekrem Duman
DEPARTMENT
Industrial Engineering
CONTACT
[email protected]
FUNDING SCHEME
British Council
START DATE
12.06.2012
2012 International Grants
DURATION
4 months
OZU BUDGET
12.196,25 GBP
54