Abstract
Chapter represents discovering behavioural patterns within non-temporal and temporal data subsets related to customer churn. Traditional approach, based on using conventional data mining techniques, is not a guarantee for discovering valuable patterns, which could be useful for decision support. Business case, as a part of the text, illustrates such type of situation, where an additional data set has been chosen for finding useful patterns. Chosen data set with temporal characteristics was the key factor after applying REFII model on it, for finding behavioural customer patterns and for understanding causes of the increasing churn trends within observed portfolio. Text gives a methodological framework for churn problem solution, from customer value calculation, to developing predictive churn model, as well as using additional data sources in a situation where conventional approaches in churn analytics do not provide enough information for qualitative decision support. Revealed knowledge was a base for better understanding of customer needs and expectations.
Top1. Introduction
Understanding of customer behaviour is a key factor of market success, especially in competitive market conditions (Bang, 2009; Berry, 1997; Berry, 2000; Giudici, 2003).
Extracting important behavioural information from transactional customer data, and enabling better decision-making throughout an organization is one of the aims when a company wants to understand their customers (Hemalatha, 2012).
New era of big data and social networks contributes in complexity of data sources for analytical purposes, and offers new challenges and also additional useful information for understanding customer behaviour (Scot, 2012; Raine, 2012). That leads us to taking in account social network analysis as an important factor for understanding hidden relations within a portfolio.
One of the frequent topics related to customer behaviour, which is important for business, is a problem of churn detection and churn mitigation (Hadden, 2006; Rashid, 2010).
Churn detection and nature of churn is closely related to customer behaviour patterns. Understanding of customer behaviour, leads us to finding solutions for successful churn detection and mitigation.
Telecom industry is very interested in churn issues. Because of that fact, the presented case study, which connects consumer behaviour, churn detection and mitigation will be from the telecom industry.
Churn detection and mitigation is also a frequent topic in data mining literature (Berry, 1997; Berry, 2000; Giudici, 2003). Telecommunication companies are also interested in churn problem solving, especially in dynamic market environment (Klepac, 2006). Customer acquisition is important, but only as a starting point of each customer lifetime cycle. Companies attempt to extend customer lifetime period as long as possible in order to return initial costs and to make profit.
Production control, planning, and scheduling are forms of decision making, which play a crucial role in manufacturing industries. In the current competitive environment, effective decision-making has become a necessity for survival in the marketplace. (Elamvazuthi, 2012)
Telecom companies are no exception, and they also use advanced analytical models for better decision making in everyday business.
With the evolution of wireless technologies, mobile networks can provide much more interesting services and resources to users than before. Consequently storing, sharing and delivering resources efficiently have become popular topics in the field of mobile networks. (Feng, 2009)
Key Terms in this Chapter
Cox Regression: One of the method for survival analysis.
Scoring: Process of assignation of some usually numeric value as grade which show as performance of observed case/object.
Churn: Interruption of the contract or using product or services.
Keywords: Data Mining, Fuzzy logic, Fuzzy expert system, Scoring, Churn, Survival analysis, Cox regression, SNA, Segmentation, Customer relationship management.
Survival Analysis: Analysis that shows survival rate (example: from population of customers) in defined period of time.
Data Mining: Discovering hidden useful knowledge in large amount of data (databases).
Fuzzy Logic: Logic which presumes possible membership to more than one category with degree of membership, and which is opposite to (exact) crisp logic.
Fuzzy Expert System: Expert system based on fuzzy logic.