Reasearch On The Online Consumer Preference And Recommendation Of Underpants Based On Data Mining

With the development of electronic commerce, businesses provide a massive commodity for consumers through the e-commerce platform, but consumers can not find the goods suitable for themselves timely and accurate, so it’s need to explore the consumer preferences, and recommend to consumers some goods may be satisfied to their preferences. This article will take the underpants as an example, the influence factor of the online consumption preference of the underpants is studied, and carry on the recommendation study according to the preference factor.Based on the previous research results, this paper puts forward a number of factors which affecting the online consumer’s preference of underpants, and divides the factors into 4 dimensions, such as network conditions, perceived risk factors, consumer characteristic factors and underpants characteristic factors. Then, through the method of statistical analysis of the sample data, the core indicator of each dimension was extracted.The results are: website interaction, website design, website convenience, process risk,visibility, user reputation, consumer personality, consumer online experience, consumer cognizance of products,the underpant’s brand, price, fabric, style, waist, pattern, type,function, color and package, etc., which set up 3 levels, 4 dimensions and 19 indicators.According to the index system, the theoretical model of the online consumer preference of the underpants’ constructed by using the structural equation method, and the verification is carried out. The results show that the influence of network conditions on the online consumer preference of the underpant is not obvious, so the initial model is modified and fitting tests are carried out. According to the modified model, the perceived risk factors, consumer characteristic factors and the product characteristic factors are significantly correlated with the online consumer preference of the underpant, but the perceived risk factor is the most significant. In the perceived risk factors, the process risk,visibility and reputation have a significant positive impact on perceived risk factors, while the process risk has the highest impact on perceived risk factors. For the consumer characteristic factors, consumer preference, online shopping experience and product cognition have significant positive influence on consumer characteristic factors, while the consumer preference has the highest impact on it. The ten indicators of underpants product characteristic factors all have significant positive effect on the online consumer preference of underpants. Underpant fabric is the most important factor in influencing factors, so it can be used as a decision attribute of the preference of the underpants.Based on the results of the research on the consumer preference of the underpants, the recommendation model of the underpant’s fabric preference is constructed by using the combination of Johnson Attribute Reduction and ID3 Algorithm. The results show that18-25 years old male tends to blended elastic fabric underpants, 26-35 years old male who likes antibacterial deodorant function tends to cotton underpants, who likes moisture absorption and perspiration function tends to blended elastic underpants, 18-25 years old women who likes moisture absorption and perspiration function tends to bamboo fiber underpants, who likes antibacterial deodorant tends to cotton underpants. 26-35 years old women who likes chic and elegant style tends to modal underpants, who likes simple style tends to cotton underwear. In order to verify the accuracy of the proposed model, the proposed model is used to verify the model. The results show that the accuracy of the model is 81.03%, and the forecasting ability is good. Finally, according to the results of the research, some suggestions are carried out.