A Knowledge-Based System for New Product Portfolio Selection

Abstract

This chapter is concerned with designing and developing a knowledge-based system for evaluating concepts of new products and selecting product portfolio. The model of measuring the product success includes metrics identified by an expert, such as duration and cost of product development or net profit from a product. The model contains a set of decision variables, their domains, and the constraints that can be described in terms of a constraint satisfaction problem (CSP). Knowledge base is specified according to CSP framework and it reflects the company’s resources, performance metrics, and relationships identified. The presented knowledge discovery process consists of the stages such as data selection, data preprocessing, and data mining in the context of an enterprise system database. In order to identify the patterns, fuzzy neural networks have been used and compared with the results from artificial neural networks and linear regression. The illustrative example presents the use of fuzzy neural networks to the identification of patterns that are translated into rules understandable by users. The proposed knowledge-based system helps the managers in selecting the most promising product portfolio and reducing the risk of unsuccessful product development.