Designing an intelligent decision support system for effective negotiation pricing: A systematic and learning approach

Abstract

Automatic negotiation pricing and differential pricing aim to provide different customers with products/services that adequately meet their requirements at the “right” price. This often takes place with the purchase of expensive products/services and in the business-to-business context. Effective negotiation pricing can help enhance a company's profitability, balance supply and demand, and improve the customer satisfaction. However, determining the “right” price is a rather complex decision-making problem that puzzles pricing managers, as it needs to consider information from many constituents of the purchase channel. To further advance this line of research, this study proposes a systematic and learning approach that consists of three different types of fuzzy systems (FSs) to provide intelligent decision support for negotiation pricing. More specifically, the three FSs include: 1) a standard FS, which is a typical multiple inputs and single output FS that forms a mathematical mapping from the input space to the output space; 2) an SFS-SISOM, which is a linear fuzzy inference model with a single input and a single output module; and 3) a hierarchical FS, which consists of several FSs in a hierarchical manner to perform fuzzy inference. To address the existing problem of a standard FS suffering from the high-dimensional problem with a large number of influential factors, a generalized type of FS (named hierarchical FS), including its mathematical models and suitability for tackling the negotiation pricing problem, is introduced. In particular, a proof-of-concept prototype system that integrates these three FSs is also developed and presented. From a system design perspective, this artifact provides immense potential and flexibility for end users to choose the most suitable model for the given problem. The utility and effectiveness of this proposed system is illustrated and examined by three experimental datasets that vary from dimensionality and data coverage. Moreover, the performances of three different approaches are compared and discussed with respect to some important properties of decision support systems (DSSs).