Intelligent e-services and multi-agent systems for B2C e-commerce

Citation

Martinez-Lopez, L. and Martinez-Lopez, F.J. (2010), "Intelligent e-services and multi-agent systems for B2C e-commerce", Internet Research, Vol. 20 No. 3. https://doi.org/10.1108/intr.2010.17220caa.001

Intelligent e-services and multi-agent systems for B2C e-commerce

Article Type: Guest editorial From: Internet Research, Volume 20, Issue 3

One of the main problems consumers face when shopping is the increasing amount of commercial information to process. It seems that more choice is not always appreciated by customers, some refer to this as the paradox of choice. It is reasonable to assume that a consumer values such a consumption scenario that offers more choice related to his/her shopping goals. But beyond a certain point information saturates the consumer’s perceptual capacity, so he/she experiences an information overload that may lead to frustration with consumption processes.

So firms must work out how to provide a balanced consumption environment where consumers feel their informational requirements are met without the confusion caused by an excessive flow of commercial information. This is especially important in Internet-based e-commerce environments, as the potential amount of information that a consumer receives during a navigational process may be huge. In particular, for new long tail-based business models like Amazon, for instance, this is a strategic issue that needs careful attention.

Online companies are personalizing their interactions with customers more and more. But, personalization needs to be effective and valued by customers. Optimally, this would involve firms working on strengthening two pillars: accurate and dynamic online customization systems; and, even more importantly, systems that empower customers to collaborate during the customization process, so they feel they are co-creators of value associated to such a process.

Different e-services related to the range of intelligent e-commerce agents have arisen in the recent years to help individuals with their online consumption processes: recommender systems, shopping systems, on-line decision support systems, to name but a few. Likewise, it is remarkable to note the integration of these systems within a general architecture based on multi-agent systems in order to attend to diversity of goals related to online consumption processes.

Both intelligent e-services and multi-agent systems have been widely analysed in computer science and machine-learning literature. They remain hot topics due to their ability to deal with different types of problems. Moreover, these are new research topics for marketing; research is still in its infancy but shows considerable potential in the short term. Especially interesting are studies that try to tackle these questions from a hybrid perspective – i.e. artificial intelligence and marketing – analyzing the effects that these systems have on the individual’s online consumption processes.

This special issue is devoted to intelligent e-services applied to B2C e-commerce. Most of the articles in this issue include extended versions of selected papers originally presented at a special session during the Twenty-eighth North American Fuzzy Information Processing Society Annual Conference (NAFIPS), held in Cincinnati (Ohio, USA) on June 14-17, 2009; a guest contribution comes in the form of paper by Kyle B. Murray, Jianping Liang and Gerald Haübl.

In sum, this special issue has seven outstanding articles treating diverse aspects of recommendation systems and intelligence-based agents applied to support B2C e-commerce. These contributions approach the general research theme from a wide range of orientations, and represent a very interesting and timely collection of papers.

In the first article, Kyle B. Murray, Jianping Liang and Gerald Haübl reflect on the necessary evolution of the Assistive Consumer Technology (ACT) habitually used to aid customers’ online decision processes. The authors describe the main factors characterizing the first generation of ACT, focused exclusively on providing high-quality recommendations to customers. They evaluate its shortcomings relative to today’s performance (check this). A variety of issues regarding ACT improvement are discussed with regard to adapting the facilitating capabilities of these technologies to the current characteristics of the web. In essence, a similar logic to the web evolution (from Web 1.0 to Web 2.0) is demanded of ACT, evolving from what the authors call ACT 1.0 to ACT 2.0.

The second article, “A multi-agent architecture to support B2C e-marketplaces: the e-ZOCO case study”, by José J. Castro-Sánchez, Raúl Miguel, David Vallejo and Vanesa Herrera, presents an architecture based on intelligent agents to facilitate user interaction in e-Marketplaces. This application searches out the most appropriate products according to requirements by means of an intelligent search agent, which can work with vague and uncertain values to define search constraints. This architecture offers advantages that are different to others in the field, in relation to scalability, efficiency, portability, safety and usability.

In “A multi-agent approach for the provisioning of e-services in u-commerce environments”, Nayat Sánchez-Pi and José M. Molina present a multi-agent system (MAS) designed for u-commerce domains. They analyze the capacity of trust management techniques in this environment and revise evaluation methods to show the benefits of context information in the use of e-services. The lack of an assessment framework for these systems leads the authors to propose a proper and suitable process for such systems, which also contains a feedback phase to take into account customer’s experiences.

The fourth article, “Using incomplete linguistic preference relations to cold start recommendations”, by Rosa M. Rodríguez, Macarena Espinilla, Pedro J. Sánchez and Luis Martínez, focuses on the cold start problem that affects collaborative filtering and content-based recommender systems. To tackle this problem adequately, they propose a knowledge-based recommendation model that treats incomplete linguistic preference relations. It requires a minimum amount of information, to be completed by an algorithm based on additive transitivity, in order to obtain useful information for the recommendations in cold start situations. This knowledge-based model is then hybridized by commutation with a collaborative recommendation model for a restaurant recommender system.

“Psychological elements explaining the consumer’s adoption and use of a web site recommendation system. A theoretical framework proposal”, by Francisco J. Martínez-López, Inma Rodríguez-Ardura, Juan Carlos Gázquez-Abad, Manuel Sánchez-Franco and Claudia C. Cabal-Cruz, presents and discusses theoretically a conceptual model which aims to offer an integrative view of the main psychological factors behind the adoption and use of commercial web site recommender agents. This contribution, with the support of an extensive literature review, highlights the adaptation and joint consideration of diverse general consumption theories and specific approaches in order to explain the acceptance of information technologies. Finally, detailed practical suggestions for managers are presented.

The sixth paper, “BizSeeker: A hybrid semantic recommendation system for personalized Government-to-Business e-Services”, by Jie Lu, Qusai Shambour, Yisi Xu, Qing Lin, Guangquan Zhang, focuses on the problem of personalizing government e-services by using recommender systems, noting that the use of collaborative filtering models are insufficient for obtaining optimum results. They propose the use of semantic information when considering features of the recommended items in order to improve recommendations. Their main contribution is a hybrid recommender system, which uses collaborative filtering, and a semantic-based model. This proposal is implemented through Bizseeker. This system helps government agencies to effectively recommend the right business partners to individual businesses, based on their requirements, interests and product categories.

The final paper, “Privacy-preserving data-mining through microaggregation for web-based e-commerce”, by Guillermo Navarro-Arribas and Vicent Torra, deals with privacy, a controversial problem in e-commerce. They focus on the use of web server logs, key data sources in e-commerce sites, and present a method to anonymize such data in order to be outsourced for marketing proposes. Their proposal consists of using statistical disclosure control techniques, in particular the microaggregation technique, which permits an acceptable degree of privacy. An application with real data is provided.

We hope the contributions to this special issue are of value to scholars and practitioners interested in the research field of intelligent agents applied to assist consumers, companies or citizens in their electronic interchanges. We would like to express our gratitude to all the authors who have submitted manuscripts for this special issue, as well as to the ad-hoc reviewers who have collaborated with us. Special thanks go to the Organizing Committee of the NAFIPS 2009 Conference for its support and collaboration in the special session linked to this issue, which we chaired there. Last, but not least, our deepest thanks to the Editor-in-Chief of Internet Research, David Schwartz, for his encouragement and unstinting cooperation with this project.

Luis Martínez-López University of Jaén, Jaén, Spain

Francisco J. Martínez-LópezUniversity of Granada, Granada, Spain