Study Of Consumer Preference Based On Data Mining Technology

In the market economy, the consumer plays an important role in contributing to the enterprise profit and social economy, that accurate measurement analysis of consumer preference should be built so that other enterprises can learn from the case and will further boost consumption. This paper, Study of Consumer Preference Based on Data Mining Technology, develops on the summaries of studies on consumer preference by other academics and the main unsupervised data mining technologies. Considering the amount of data used in the paper is over 5 million and involved several large sheets, this paper specially states the basic preparation work of data, including the source of data, data importing, data merging and data cleaning as well as dealing with the missing value. Besides, descriptive statistics is done from the perspective of shops, goods, and bills, so that big date can be shown visually. Based on the theories of preference construction and consumer behavior, and combining with product life circle and improved Bass model, the paper constructs a dynamic model of consumer preference measurement, which transfers the preference from an incalculable variable to a calculable one. In the case study, consumer preference is represented by the sales, and is computed to show the difference of consumer preference in various goods using the model constructed above. Association rules are adapted to the mode of consumer purchase. And the rules of colors, types of goods and shops which consumer prefer are found. Finally, the paper summarizes that data mining using in the consumer preference study based on data library can classify the consumer and dig out the pattern of purchase, the applicability of the model, a dynamic model of consumer preference measurement, is wide.