Editor’s Introduction 14(2)

Vladimir Zwass
International Journal of Electronic Commerce,
Volume 14 Number 2, Winter 2009-10, pp. 5.

Abstract: Having become a fundamental global infrastructure, the Internet brings vast opportunities–and risks. Among the vulnerabilities on various levels of the Internet stack, flaws in software security loom large. It has been established by both research and practice that software vendors have an incentive to release first and patch later. This leads, of course, to multiple vulnerabilities out of the virtual box. The problem is beginning to be addressed in the legal domain. This is important in any case. However, the design of incentive-compatible economic mechanisms appears to be a far better avenue to an effective solution. Byung Cho Kim, Pei yu Chen, and Tridas Mukhopadhyay present and analyze here a mechanism that aims at prevention by giving vendors an incentive to offer new products with en-hanced security. In this scheme, the risk due to security exposures latent in the software would be shared between vendor and customers. The authors find that voluntary risk-sharing can be attractive to new market entrants under certain conditions. This innovative market-driven solution to a major problem will certainly be further developed by future research.

Jean-Claude Usunier, Nicolas Roulin, and Björn Sven Ivens present the results of an encompassing study of B2B Web sites. They empirically relate the design and content of these sites to the multiple dimensions of recog-nized cultural and industry differences. The authors find indeed a substan-tial influence of cultural factors. Their analysis shows that some of these influences may be counterproductive to corporate aims. Further, the authors observe the underuse of the potential of e commerce. By the very definition of e commerce, Web sites have two fundamental functions: to provide business information and to facilitate business transactions. It is notable that the authors find a predominance of the first use of Web sites at this quite advanced e commerce stage. In some cases, this is the reasonable consequence of seeking a relational approach with potential or actual cus-tomers. In others, it calls for a review of the corporate approach to e commerce, just as the cultural limitations discovered by the authors call for a review of corporate Web site content and design.

Predictive recommender technologies have by now become a hallmark feature of e commerce. They help sellers to bring attractive products to the attention of appropriate customers, whom they also help to handle the cog-nitive complexity of product selection. The collaborative filtering approach to recommendation does not require explicit construction of customer profiles and relies on the past behavior of the customer and of customers considered similar. A vast, sparse matrix of customers and of products they have rated conceptually represents the database for the algorithms. Since the matrix does not by any means contain most of the potential customer-product ratings, overlapping of ratings between customers in order to make recommendations based on the ‘customers like you’ principle requires estimation of the very many missing ratings. Carme Julià, Angel D. Sappa, Felipe Lumbreras, Joan Serrat, and Antonio López present a novel method for predicting the missing ratings at a lower computational cost than the alternatives. The authors show superior results in several experiments that are accomplished at a lower computational cost than the alternatives.

The functioning of recommenders is just one consideration for preventing Web site account sharing. Lost revenue from customers or advertisers is among the consequences. Detection of sharing is not easy, and intrusive methods can deflect many users/customers. Seong seob Hwang, Hyoung joo Lee, and Sungzoon Cho present a detection method based on keystroke dynamics. The keystroke pattern has been long recognized as an individual characteristic, among other biometrics. The authors offer a generic method of preventing shared use, based on any form of clustering a user’s keystroke data. The proof-of-concept cases show the promise of the method.

Interoperability of enterprises to be integrated in a supply chain is understood to mean the ability of firms and their information systems to work together relatively seamlessly. Integration experiences show that ‘relatively’ is the operative word here, and that interoperability does not come naturally. There is value in unpacking different aspects of this capability. In the final paper of the issue, Spiros Mouzakitis, Aikaterini Maria Sourouni, and Dimitris Askounis consider a hierarchical model of interoperability, from the business to the network layer. They empirically study the relationship between the presence of interoperability on each of these levels and the required integration effort. The authors find that interoperability at the business level, which may be considered strategic alignment of the en-terprises, has the most unequivocal effect on reducing integration effort. They also offer a nuanced analysis of the effects of the interoperability or its absence on the other four hierarchical levels. The results will help in esti-mating the effort required to bring enterprises together in supply chains and further to assess the advisability of such an effort. It may be extended in the future to compatibility considerations in mergers and acquisitions.