Analysing Incomplete Consumer Web Data Using the Classification and Ranking Belief Simplex (Probabilistic Reasoning and Evolutionary Computation)

Consumer attitudes, involvement and motives have long been identified as important determinates of decision making in classic models of consumer behaviour. Online consumer attitudes may differ depending on the level of web experience of the intended consumer. This chapter considers Classification and Ranking Belief Simplex (CaRBS) analyses of consumer web data, considering attitudes from consumers with different levels of web experience. The CaRBS technique is based on Probabilistic Reasoning (Dempster-Shafer theory) and Evolutionary Computation (Trignometric-Differential Evolution), two known components of soft computing. An important facet of the presented analyses is the ability of the CaRBS technique to analyse incomplete data, without the need for the missing values present to be managed in anyway. The chapter allows a pertinent demonstration of how soft computing, here using CaRBS, can offer the opportunity for realistic analysis, more realistic than traditional techniques.