Research On Consumer Preference Measure Based On Belief Rule Base

Consumer preference is the important factors to affect consumer behavior and consumer demand. In marketing activities, consumer preferences for new product design, advertising strategy formulation, the choice of the ways of distribution, market segmentation and positioning plays an indispensable role. It is very difficult to accurate quantitative comparison due to the subjective characteristics of preference. Accurate measurement is very important to consumer preferences to provide reliable basis for marketing decision-making. There is a certain causal relationship between consumer preference and attribute of goods and commodities, and the causal relationship is nonlinear. Belief rule-based can affect nonlinear causal relationship among attributes. Subjectivity of Expert rules is stronger, Mining association rules theory can effectively mine potential correlation in the data. In the decision-making process inevitably exists uncertainty, Evidence reasoning could properly deal with uncertainty evidence reasoning and effectively integrate the independent evidence.This paper firstly summarizes the preferences of consumers, the association rules and evidence theory. Secondly build a model to measure the belief rule base system is constructed based on consumer preferences, use the association rule mining theory, learning rules from data. Finally, choose the tea consumer preferences as a case application. Because of product variety and intense compitition in tea market, research and development of new products need accurately grasp the consumer preferences to meet the market demand. There is a causal relationship between the nonlinear properties of numerous consumer preferences and the tea attribute. Through the questionnaire investigation of the tea market consumer preferences to get the data, the measurement model is established with the help of software mining rules to measure consumer preference score to the different tea. Contrast discovers that caculation results by belief rule-base are basically identical with the actual result, which prove that rule-base system can provide objective and accurate method for market preference measure to offer important decision-making information for product development.