EDITORIAL
Web mining applications in E-commerce and E-services
I-Hsien Ting
Department of Information Management, National University of Kaohsiung,
Kaohsiung, Taiwan
Abstract
Purpose – The purpose of this guest editorial is to introduce the papers in this special issue. Design/methodology/approach – A brief introduction about the issue of Web-mining applications in e-commerce and e-services is provided, along with a summary of the main contributions of the papers that are included in the special issue.
Findings – The value of Web mining techniques can be enhanced through applying them to real environments such as e-commerce and e-services. The research fields of Web mining, e-commerce and e-services can also be expanded.
Originality/value – An overview of the special issue and related research is provided in this paper. Keywords – Web mining, e-commerce, e-services, data mining, Web intelligence.
Paper type – General review.
From Web-mining techniques to applications
In recent years, Web mining has become a popular area of research, integrating the different research areas of data mining and the World Wide Web. According to the taxonomy of Web mining proposed by Cooley et al. (1999), there are three sub-fields of Web-mining research: Web usage mining, Web content mining and Web structure mining. These three research fields cover most content (such as text, graphics, linking structure, etc.) and activities (such as navigation behavior, transactions, etc.) on the Web. With the rapid growth of the World Wide Web, Web mining has become a hot topic and is now part of the mainstream of Web research, such as Web information systems and Web intelligence.
During the last decade, many researchers have devoted their energies to Web-mining research. The aim is to apply or improve traditional data-mining techniques in order to adapt them to the new analysis target and data source. In addition, some of this research focuses upon developing new Web-mining techniques and algorithms. However, such work is meaningless if these techniques and algorithms are not applied to a real environment. Therefore, it could be the right time to shift the research focus from Web-mining techniques to applications.
Among all of the possible applications in Web research, e-commerce and e-services have been identified as important domains for Web-mining techniques (Kohavi et al., 2004). Web-mining techniques also play an important role in e-commerce and e-services, proving to be useful tools for understanding how e-commerce and e-service Web sites and services are used, enabling the provision of better services for customers and users. Thus, the special issue will focus upon Web-mining applications in e-commerce and e-services.
The possible research topics of Web-mining applications in e-commerce include customer behavior analysis, transaction analysis and Web site design. The possible research topics of Web-mining applications for e-services include e-banking, search engines, on-line auctions, on-line knowledge management, social networking, e-learning, blog analysis, and personalization and recommendation systems. Through applying Web-mining techniques to e-commerce and e-services,
value is enhanced and the research fields of Web mining, e-commerce and e-services can be expanded.
In this special issue
On July 23, 2007 I chaired an international workshop in Tokyo, Japan entitled “Web mining for e-commerce and e-services”. The workshop was held in conjunction with the IEEE joint conference on E-commerce Technology (CEC ’07) and Enterprise computing, E-commerce and E-services (EEE ’07). After the workshop, the OIR agreed to publish a special issue of the journal to address this research topic. Extended versions of the best papers featured in the workshop have therefore been selected for inclusion in this special issue. We also made a public announcement calling for further high quality papers, some of which have been accepted for inclusion in this special issue of OIR.
The first paper is Rong Gu, Miaoliang Zhu, Liying Zhao and Ningning Zhang’s “Interest mining in a virtual learning environment”, which uses behavior analysis to analyze interest mining in e-learning. A prototype system has been developed and proposed in their paper in order to proceed with experimentation using their interest mining approach. They also present their research results concerning the concept classification of behavior in VLE, and a hierarchical interaction model is proposed based on the classification.
On-line auctions are very popular Web applications, and the second paper “An application of Web-based data mining: Selling strategies for on-line auctions” by Yanbin Tu is an interesting discussion of the use of Web mining in order to develop better selling strategies for on-line auctions. Tu developed a Web data-mining system to collect data from eBay, applying the CART (classification and regression tree) technique to analyze the collected data in order to generate effective selling strategies for on-line auctions.
Another popular Web application is P2P. Shuai Yuan and Jun Shen contributed the third paper, entitled “Mining e-services in P2P-based workflow enactments”. The authors proposed a set of QoS (quality of services) specifications and mechanisms to monitor P2P workflow, derived from observing and analyzing the changes in e-services workflow. The proposed QoS-OWL approach helps to dynamically select an appropriate oriented peer in a P2P network so that service-oriented business process can be enhanced.
Georgios Lappas provides a survey of Web-mining research related to areas of social benefit in his paper “An overview of Web mining in societal benefit areas”. In this paper, Lappas firstly introduces the taxonomy of Web mining, and then the six research areas of Web mining that are of societal benefit: e-learning, e-government, e-politics and e-democracy, helpdesks and recommendation systems, digital libraries, and public security and crime investigation.
Yingzi Jin, Yutaka Matsuo and Mitsuru Ishizuka have contributed a paper entitled “Extracting inter-firm networks from the World Wide Web using a general-purpose search engine”, in which the important research topic of social networking is addressed. In this paper, a novel approach is proposed in order to extract the inter-firm social network using an enhanced search engine and text-processing algorithm. The algorithm uses relation keywords to characterize the names of firms that co-appear coincidentally on the Web in response to a query. The authors also present an empirical study to show how the proposed approach works and to evaluate the approach and algorithm.
The sixth paper is a conceptual paper entitled “Perceptual navigation in Web mining” by Adel Elsayed, which discusses how Web mining can be performed by human agents, and reviews the techniques. The nature of artefacts is introduced and four types of artefacts are identified: cognitive externalizations, artistic expressions, communicative accounts, and factual records. The author then introduces the concept of information space and how to map text into information space, as well as how information space should be constructed.
insurance companies” proposes a model that uses data-mining techniques for business rule generation and recommendation. Two data-mining techniques – C4.5 and apriori algorithm – are introduced in the paper to analze profile and browsing logs. The authors propose that this model can be used to increase the application rate for policy loans, thereby increasing interest revenue for any insurance company that adopts the system. The proposed approach has also been implemented as a system, and real data have been used for their case study.
“A rapid egocentric search scheme using authority estimation in blog space” is contributed by Yoonjae Jeong and Dongman Lee, covering a new fashion media-blog. The authors propose a rapid egocentric search scheme to search documents in a linked (neighboring) blog. The scheme can narrow down the search space to more important blogs by predicting the authority scores on the basis of the local information of the blog. Performance evaluation also shows the speed of the egocentric search process and the quality of the retrieved documents.
The final paper in this special issue addresses the application of Web mining to e-learning, entitled “eLORM: Learning object relationship mining-based repository” by Yang Ouyang and Miaoliang Zhu. The term “eLORM” means learning object relation mining, which is an interesting way of using mining techniques to analyze the learning object relation based on the learner’s usage information in an LO repository. The analyzed results are then used to recommend various learning objects to learners. The authors also propose system architecture to show how the proposed approach can be used to discover and analyze learning object relationship patterns.
Reference
Cooley, R., Mobasher, B., and Srivastava, J. (1999) “Data Preparation for Mining World Wide Web Browsing Patterns” Knowledge and Information Systems, Vol. 1 No. 1, pp. 5-32
Kohavi, R., Mason, L., Parekh, R., Zheng, Z. (2004) “Lessons and Challenges from Mining Retail E-commerce Data” Machine Learning, Vol. 57 No. 1-2, pp. 83-113
