An e-customer behavior model with online analytical mining for internet marketing planning
Introduction
In e-commerce, the current challenge is determining how to design responsive Web site infrastructure that provides a sustainable competitive advantage through a better understanding of target customers. The quality of an e-commerce site depends on interrelated factors such as site architecture, network capacity, Web services, and the unpredictability of e-customer behavior. These characteristics imply the need to measure the behavior of the Web-based system and its users. Knowledge management is the key to business learning. The technologies that support knowledge management in e-business are data warehousing, data mining, the Internet, and document management systems [21], [25], [26].
Online marketing aims to produce online revenue by understanding customer needs. Meeting this objective requires knowledge of how e-customers' online movements change from awareness of products to the exploration of options and further to purchase commitment. An online analytical mining (OLAM) system using an underlying cognitive model and e-customer behavioral graph can be used to articulate the online activities of e-customers on a particular Web site. This can provide the framework of an e-customer behavior (eCB) model that can be used to discover e-customer profiles which identify the significant dimensions of online behavior and identify Web pages that trigger behavior changes. The knowledge thereby obtained will foster informed Internet marketing decision making, and allow Web content and infrastructure refinement to support Internet marketing.
Section snippets
Current background, theoretical underpinnings and hypotheses
Electronic commerce (EC) is growing rapidly, and offers a diversity of related issues to investigate. Ngai [18] presents a literature review and classification scheme for EC research. Over 78% of EC research has been focused on applications, implementation and technical issues, and only 9% has touched the topic of e-customers, with very few studies directly addressing the issue of e-customer preferences and their effects on Web site acceptability. Because e-customers learn fast and want Web
The OLAM methodology
Our OLAM system for path traversal patterns includes incremental Web usage mining updates [8], [13], [27], [28]. It stores the derived Web user access paths in a data warehouse. The system updates the user access path pattern in the data warehouse by data operation functions that are automated by webmaster. The result is an OLAM that uses the underlying e-customer behavior graph, which is capable of discovering association semantics between tick sequences, e-customer profiles for customer
Prototype
We verify the eCB model by using a prototype. Section 4.1 provides an outline of our OLAM process with data flow diagrams. Section 4.2 explains the prototype system. The Web log file was collected from the Computer Science Laboratory Web site of the City University of Hong Kong. The site hosts a variety of information, ranging from departmental information and courses to individual Web sites. We identified seven pages for our empirical study.
- Page 1 (P1):
Department history, facilities, and message from
Conclusions and future research
This paper proposes an eCB model that uses an OLAM methodology to discover e-customer behavioral changes on a Web site to support Internet marketing. We undertake an empirical study using a prototype built upon the eCB model to verify its feasibility. Our prototype identifies different customer segments according to their successful path frequencies, counting sessions that result in purchases, the referring Web pages that determine targeted e-customer behavioral changes, and, based on the
Acknowledgments
The authors acknowledge the valuable and constructive comments that were received from the editor and the three referees. We are indebted to them for their invaluable advice, which has greatly improved the quality of this paper.
This work was greatly supported by the University Grant Committee of Lingnan University of Hong Kong (grant code: DR03A5); Irene Kwan, Department of Computing and Decision Sciences, Lingnan University, Hong Kong.
Dr. Irene S.Y. Kwan is an associate professor of the Computing and Decisions Sciences Department at Lingnan University of Hong Kong. She is also the Assistant Director of business programmes in her University. She received her PhD degree from Brunel University of London in 1999. Kwan has published over 30 research papers in key journals and conferences. Her research interests are in intelligence business, knowledge management, knowledge discovery and data mining.
References (28)
- et al.
Characterizing browsing behaviors on the world wide web
Computer Networks and ISDN Systems
(1995) - et al.
Marketing on the Internet—who can benefit from an online marketing approach?
Decision Support Systems and Electronic Commerce
(2000) Complementing search engines with online web mining agents
Decision Support Systems
(2003)- et al.
A literature review and classification of electronic commerce research
Information and Management
(2002) - et al.
Dynamic rule refinement in knowledge-based data mining systems
Decision Support Systems
(2001) - et al.
Knowledge management and data mining for marketing
Decision Support Systems
(2001) - et al.
Data Warehousing in the Real World
- et al.
Denotation and connotation in human–computer interface: the ‘save as…’ command
Behaviour and Information Technology
(2004) Create customer-effective e-services
e-Business Advisor, Technology Strategies for Business Innovators
(2000)- et al.
Antecedents of B2C channel satisfaction and preference: validating e-commerce metrics
Information Systems Research
(2002)
Web mining for web personalization
ACM Transactions on Internet Technology
Data Mining Concepts and Techniques
Scalable parallel data mining for association rules
ACM
A Cube Model For Web Access Sessions And Cluster Analysis
Cited by (60)
From free to fee: Monetizing digital content through expected utility-based recommender systems
2022, Information and ManagementCitation Excerpt :The ARM approach aims specifically to gauge the extent to which the sale of one item influences the sale of another, determining the relational strength of the two items based on the frequency with which they are sold simultaneously in one transaction. The ARM method has been widely applied and utilized as a key analytical tool in numerous business contexts, including marketing promotion, cross-selling [6, 17, 28], website analysis [5, 52, 60], decision support [31, 41, 65], credit evaluation [15], privacy issues [38, 54], criminal event prediction [71], customer behavior analysis [32], and fraud detection [39]. Despite its wide application in practice, ARM is often criticized for its dependence on the frequency of transactions and strength of affinity between items while paying scant attention to the business significance and impact of quantitative patterns [21, 72].
Using Data Sciences in Digital Marketing: Framework, methods, and performance metrics
2021, Journal of Innovation and KnowledgeDevelopment of conceptual model to increase customer interest using recommendation system in e-commerce
2021, Procedia Computer ScienceTime-Dependent Enhanced Personalized Semantic Map-Based Transportation Mode Recommendation System
2024, Lecture Notes in Networks and SystemsMarketing Automation for Digital Transformation: A Review
2023, AIP Conference ProceedingsReplenishment and delivery optimization for unmanned vending machines service system based on fuzzy clustering
2023, Electronic Commerce Research
Dr. Irene S.Y. Kwan is an associate professor of the Computing and Decisions Sciences Department at Lingnan University of Hong Kong. She is also the Assistant Director of business programmes in her University. She received her PhD degree from Brunel University of London in 1999. Kwan has published over 30 research papers in key journals and conferences. Her research interests are in intelligence business, knowledge management, knowledge discovery and data mining.
Dr. Fong is an associate professor of the Computer Science Department at City University of Hong Kong. He received his PhD degree from University of Sunderland in 1993. Fong has worked in the electronic data processing in US for more than 10 years. His industrial expertise and research interests are in database, data warehousing, data mining and XML. He has published more than 30 academic journal papers in these areas.
Hing Kwok Wong is a PhD student in the Computer Science Department of City University of Hong Kong. He received a BSc in Information Technology, with First Class Honors, in 1999 and a M.Phil degree in Computer Science, in 2001 at City University of Hong Kong. He has published several papers in journals and conferences. His current research interests are online analytical mining, web usage mining, and XML-enabled database.