Integrated view of business-to-business electronic commerce based on intelligent agent and data mining

Abstract

The explosive growth of the Internet and the rapid adoption of the Internet as a commercial medium have changed fundamentally the way business-to-consumer (B2C) and business-to-business (B2B) commerce is conducted in the world. They have generated significant interest in the development of B2C and B2B markets all over the world.

Nowadays the maturity of the B2C markets has resulted in two rigorous research areas, such as agent mediation and e-enabling customer relationship management (CRM). As well as in B2C electronic commerce (EC) environment, agent technologies and CRM play important roles in B2B EC markets to facilitate business transactions and increase company’s revenue, and to gain competitive advantage.

The rigorous researches on the agent technology are mainly due to the existence of the non-automated buying process, from which can be more or less relieved by using software agent technologies. The researches on CRM are fundamentally attributable to the increasing competitiveness in the retail business. In such a competitive environment a successful retailer must provide a bundle of customized services in order to survive and gain competitive advantage. These are why this thesis considers them as the important themes and thus focuses on the development of e-enabling static and dynamic CRM models, which can be utilized in B2C or B2B commerce.

This thesis introduces the static CRM model and proposes the new one-to-one marketing strategies utilizing it in B2C EC environment. In order to build successful e-CRM solutions in the B2C markets, this thesis utilizes data mining techniques. One-to-one marketing strategies utilizing the static CRM model allow retailers to analyze the offline and online data about customers in order to capture their buying behavior patterns. They also allow retailers to select the right customers on whom to focus, and to offer the right marketing campaigns including direct mails and e-mails to the existing customers, and ...