Green Product Retrieval and Recommendations System

Green Product Retrieval and Recommendations System

Yi-Chun Liao
DOI: 10.4018/978-1-60566-114-8.ch017
(Individual Chapters)
No Current Special Offers


This chapter introduces a preference-based recommendation procedure in a green product information retrieval system. It constructs a green information management system based on data mining technology. Green products and relevant green regulations were collected and then integrated in a content-based and collaborative filtering method to provide a preference-based query interface for green products. It is hoped that the proposed system offers consumers a green information platform when they are considering buying green products. Our proposed system recommends the best possible choices for consumers that indicate a green preference. Besides serving as a green information retrieval for the consumers, the system also assists the product designers with understanding the preference some consumers have for green products and the satisfaction they get from buying these products.
Chapter Preview

Management Information System

The Trend of E-business

First, business and technology trends are changing rapidly. More and more enterprises increase their portion of the e-market and develop their information systems (Post & Anderson, 2003). Utilization of the Internet has changed how people network and communicate, and the worldwide Web has changed how we obtain information.

Consequently, customer-oriented interactive systems are becoming a major trend in the development of the current e-business system (Blecker et al., 2005). Information systems should support the requirements of the customers while automating the operating process, allowing customers to configure their products by specifying the attributes of the products they are looking for (Bramham & MacCarthy, 2003). In order to configure a system for a user, the system requires an accurate understanding of the customer’s needs so as to create a complete description of a product that suits the consumer’s individual requirements. Given a set of customer requirements and a product family description, the task of configuration is to find a valid and completely specified product among all of the alternatives (Sabin & Weigel, 1998). Up to now, the product configuration process has been a very technical-oriented process, necessitating product expertise of the customer while seldom taking into account the requirements of the customer.

When it comes to configuration knowledge, there are three important design approaches based on: a) rule-based, b) model-based, or c) case-based, respectively (Sabin & Weigel, 1998). The rule-based knowledge representation method relies on rules which have the following form: “if condition then consequence,” which is the most common one implemented in practice. The main assumption behind model-based reasoning is that the system’s model consists of decomposable entities and interactions between their elements. For configuration problems with high product complexity, model-based approaches are more convenient than rule-based approaches. The case-based approach relies on the assumption that similar problems have similar solutions. The knowledge necessary for this type of reasoning consists of cases that record a set of configurations sold earlier to customers. The implementation of a case-based reasoning system is appropriate when the solution space consists of only a few products.

Furthermore, suppliers never stop improving the advisory quality in mass customization. Muther (2000) provided a supplier-customer model that structures the relationship between customers and suppliers in four phases: stimulation, evaluation, purchasing, and after-sales. A customer advisory is especially important during the evaluation phase in which the customer evaluates different product offers in order to decide which product to buy. In this phase, the customer must process the information about the different products that are available. To ensure a high level of customer satisfaction as well as an optimal personalized advisory, valuable data about the customers must be obtained for all of the four phases, which has to be automated and offered as an e-service today.

In the existing customer advisory systems, recommender systems have been successfully implemented in online shops. Generally, they support people who have little or no product knowledge in making a suitable choice (Resnick & Varian, 1997).

Complete Chapter List

Search this Book: