A Dominance-based Rough Set Approach to customer behavior in the airline market
Introduction
In the face of a highly competitive and fast-changing airline market, managers must not only provide high-quality service but also react appropriately to changes in customer needs. However, it would be helpful if, instead of targeting all customers equally or offering the same incentives to all customers, enterprises could target only those customers who meet certain profitability criteria based on their individual needs or purchasing behaviors [5]. Customer behavior is the result of complex interactions between a number of factors, which can include the level of marketing activity, the competitiveness of the environment, brand perception, the influence of new technologies, and individual needs [34]. The characteristics and behaviors of airline customers are even more complex, and customer perception and behavior are affected by many factors, such as safety, service, technology, environment, price and many others. Hence, it is crucial that management determine the most important factors that affect the attitude and loyalty of airline customers. In the past, researchers have generally made use of statistical surveys to determine customer behavior. In such surveys, natural language or linguistic variables (e.g., “although the airline service is satisfactory, the price of the product being offered is high, the individual’s decision is not to purchase”) are used to describe customer patterns. Unfortunately, this can create an environment of imprecision, uncertainty, and partiality with regard to knowledge. These linguistic variables are then transformed into quantitative values, after which factor, cluster, and discriminant analyses are conducted. However, the semantic imprecision of natural languages leads to problems of computation, especially when the information described in a natural language is beyond the reach of existing bivalent logic and probability theory techniques [41].
Recently, data mining techniques have been adopted to predict customer behavior [6], [30]. Data mining is one stage in Knowledge Discovery in Databases (KDD), involving the application of specific algorithms for pattern extraction [21]. Marketing managers can develop strategies to attract new customers and retain highly valued ones based on this mined knowledge. The Dominance-based Rough Set Approach (DRSA), originally developed by Greco et al. [8], [9], is a relatively new approach in data mining that is very useful for data reduction in qualitative analysis. The rough set theory, a kind of natural language computation, is particularly useful for dealing with imprecise or vague concepts [23]. Basically, natural language computation is a system in which the objects of computation are simply predicates and propositions drawn from a natural language. A set of decision rules is generated by applying the rough set approach to analyze the classification data. These decision rules are in the form of logic statements of the type “if conditions, then decision”. The set of decision rules represents a preference model for the decision-maker that is expressed in a natural and understandable language. According to Zhu et al. [40], the rough set method does not require additional information about the data; it can work with imprecise values or uncertain data, is capable of discovering important facts hidden in that data, and has the capacity to express them in natural language. The rough set theory has been successfully applied in a variety of fields, including medical diagnosis, engineering reliability, expert systems, empirical studies of material data [15], evaluation of bankruptcy risk [29], machine diagnosis [39], business failure prediction [1], [4], network intrusion detection [40], travel demand analysis [7], mining stock price [35], the insurance market [28], and accident prevention [36].
Although the Classical Rough Set Approach (CRSA) is a powerful tool for handling many problems, it is not able to deal with inconsistencies originating from the criteria, e.g., attributes with preference-ordered domains (scale) like product quality, market share, and debt ratio [10]. However, the DRSA has an advantage over the CRSA in that it has access to an information table that displays comprehensive dominance relations. It is able to deal with inconsistencies where decisive classes are not consistent with their criteria. The aim of this study is to mine data regarding airline customer behavior using the DRSA. The derived knowledge can help airlines identify valuable customers, predict future behavior, and enable firms to make proactive, knowledge-driven decisions.
Section snippets
Customer behavior
In most research on customer behavior, customer demographic variables are applied to analyze customer behavior [30]. However, Rayport and Sviokla [27] suggest that a customer’s perception of product or service value is comprised of three basic elements: the product or service that a company offers, the context in which a company offers this product or service, and the infrastructure that enables the transaction to take place. In the traditional marketplace, content, context, and infrastructure
Basic concepts of the Dominance Rough Set Approach
The rough set theory, firstly introduced by Pawlak [22] in 1982, is a valuable mathematical tool for dealing with vagueness and uncertainty [23]. For a long time, the use of the rough set approach and other data mining techniques was restricted to classification problems where the preference order of the evaluations was not considered. This is due to the fact that this method cannot handle inconsistencies that occur as a result of the violation of the dominance principle [10]. In order to deal
Empirical study: a case of customer behaviors in the airline market
In order to demonstrate the effectiveness of the DRSA and our proposed approach, we carried out an empirical study that is described in this section. We produced a questionnaire about customer behavior in the airline market with single and multiple-choice answers and used DRSA to explore the classification problem. The results should provide airlines with useful information to help them develop marketing strategies and achieve their marketing goals.
Discussion
Using the same data set and dummy-variable technique, we also apply discriminant analysis. The results are shown in Table 5. Clearly the DRSA shows better prediction ability than does discriminant analysis. The hit rate increases from 91.8% for discriminant analysis to 95.6% for DRSA. Furthermore, in contrast to discriminant analysis, the rough set theory requires no underlying statistical assumptions. In particular, the DSRA can handle both attributes/criteria with and without preference
Conclusions
This study illustrates the usefulness of the DRSA approach as an operational tool for the prediction of customer behavior in the air transport market. The proposed prediction model takes the form of decision rules. Since the derived rules are supported by real examples, they describe only the most relevant attributes/criteria. The classical rough set theory handles attributes without preferences, a technique that does not always accurately represent the real world. The DRSA is constructed by
Acknowledgements
This paper was supported by National Science Committee of Taiwan under Grant No. NSC 98-2410-H-424-005.
References (41)
- et al.
Variable precision rough set theory and data discretisation: an application to corporate failure prediction
Omega
(2001) - et al.
Multi-criteria classification – a new scheme for application of dominance-based decision rules
European Journal of Operational Research
(2007) - et al.
Mining changes in customer behavior in retail marketing
Expert Systems with Applications
(2005) - et al.
Business failure prediction using rough sets
European Journal of Operational Research
(1999) - et al.
Data mining of association structures to model customer behavior
Computational Statistics and Data Analysis
(2002) - et al.
Incorporating the rough sets theory into travel demand analysis
Tourism Management
(2003) - et al.
Rough sets theory for multicriteria decision analysis
European Journal of Operational Research
(2001) - et al.
Rough sets methodology for sorting problems in presence of multiple attributes and criteria
European Journal of Operational Research
(2002) An integrated data mining and behavioral scoring model for analyzing bank customers
Expert Systems with Applications
(2004)- et al.
Rough sets applied to material data
ACTA Material
(1996)
A stochastic model of e-customer behavior
Electronic Commerce Research and Applications
Stochastic dominance-based rough set model for ordinal classification
Information Sciences
An e-customer behavior model with online analytical mining for internet marketing planning
Decision Support Systems
Predicting customer behavior in the market-space: a study of Rayport and Sviokla’s framework
Information and Management
Evolutionary modular design of rough knowledge-based network using fuzzy attributes
Neurocomputing
Rudiments of rough sets
Information Sciences
Rough sets and Boolean reasoning
Information Sciences
Rough sets: some extensions
Information Sciences
Rough set theory in analyzing the attributes of combination values for the insurance market
Expert Systems with Applications
Mining the change of customer behavior in an internet shopping mall
Expert Systems with Applications
Cited by (123)
Three-way decisions based multi-attribute decision-making with utility and loss functions
2024, European Journal of Operational ResearchA computationally efficient approximation calculation method in dominance-based rough set approach
2023, Applied Soft ComputingIdentification and severity assessment of challenges in the adoption of industry 4.0 in Indian construction industry
2023, Asia Pacific Management ReviewA parallel rule-based approach to compute rough approximations of dominance based rough set theory
2022, Engineering Applications of Artificial IntelligenceA regret theory-based three-way decision approach with three strategies
2022, Information SciencesNovel fuzzy rank discrimination measures for monotonic ordinal feature selection
2022, Knowledge-Based Systems