Access-Control Architecture to Support E-CRM and Distributed Data Mining Simon Fong, Zhuang Yan Serena Chan Faculty of Science and Technology e-Business Development Team Universidade de Macau Omron Asia Pacific Av. Padre Tomás Pereira S.J., Taipa U.E. Square Macau SAR Singapore {ccfong@umac.mo syz@umac.mo} serenachan@omron.com.sg Abstract – This paper presents an Access-Control (AC) architectural model for Distributed Data Mining (DDM), which is tailored to meet the needs of online Customer Relations Management (CRM) where the services are personalized and the web content delivery are targeted to specific segments of customers. This hybrid AC architecture integrates the client-server and the mobile agent technologies that can easily be upgraded from most existing web-based platforms. The AC model focuses on the online e-personalization (for CRM) and web intelligence issues (for decision support), which are particularly relevant in the context of making a website customer-service oriented while enabling top management to gain insights from the operation of the website. Technically, a software agent architecture that consists of functions of access-control, customer-profiling, dynamic web pages creation, data-mining and a rule-based system is employed. Also in this paper, we argue that not only we can provide richer business intelligence by profiling customers than by merely mining raw web-logs for the traffic statistics, a new paradigm of business insights and e-CRM functions can be obtained and deployed respectively, by combining the customers’ profiles and traffic statistics under this AC architecture. I. INTRODUCTION Nowadays, most companies have moved into e-business where they use Internet as a platform to interact with their customers. This has created more touch points than the traditional face-to-face visits, and therefore we have a new realm of approach called e-CRM to deal with. Similarly to the traditional CRM, e-CRM is a methodology used to build and maintain a loyal customer relationship online and to discover the most profitable way to build that relationship through e-mails, interactions at the website, and personalizing web contents for customers. It is an ever ongoing process of acquiring, retaining and growing profitable customers via the Internet. Specifically, e-CRM adheres closely with three business areas: decision support, personalized service, and customer management. One must have a very good understanding about the operation of the business as well as the customers, so that sound decisions can be made by the management on how to run a successful e-CRM. Since e-Business is like any other business perhaps at a faster pace, its business information evolves and changes continuously. Business intelligence must be obtained timely in order to remain effective. A more specific kind called Web intelligence will be discussed later on. In e-CRM, the two business areas that operate at the front-line are Personalized service and Customer management. Personalized service is about providing better customer service on the website. This is done by presenting the right information to the right customers. Users can easily access to the information that they want, receive the right piece of recommendation, etc. for example. So that would enhance the user’s experience on the website, hoping that they will like to visit the same website again. The latter business area that e-CRM covers, is in general called Customer management that operates through the touch-points which the customers communicate with the company on the Internet. Instead of passively serving the customers by customizing the web content for them to view, Customer management proactively contact them on the web via emails, e-newsletters and or an interactive help-desk. Sales force automation may also fit into Customer management. It is however beyond the scope of this paper as we focus on e-business services. Together, personalized service and customer management help retain good customers, win their loyalty, and may attract new ones by earning word-of-mouth from existing satisfied customers. Figure 1. e-CRM covers 3 business areas and their functions In general, for targeting the business areas of decision support and personalized service, we have identified three main technologies namely Data-warehousing (that includes data-mining modules and decision support modules), Web log analyzer [1] and Web spy [2]. Each one of them provides certain insight or understanding of one of the many aspects of a business respectively, as shown in Figure 2. In particular, we state that having a web log analyzer alone is not sufficient to provide a complete picture of business intelligence. Nevertheless, web-log analysis gives very good detailed reports on web site traffic statistics. Other information such as competitors, market positions, and most importantly the customers and their behaviors on the website needed to be known. In our case, e-Personalization that means personalizing web-content according to the customer’s interest/needs will be guided by the knowledge that we have on him.