Research On A Big Data-based Causal Inference Recommended Model For Online Garments Among College Students

With the development of e-commerce, there is an important problem needed to solve for e-commerce website, which is information overload. The current all kinds of personalized recommendation methods are based on the correlation between the underlying data, focusing on the breadth of the big data, failed to analyze the big data on depth, and thus did not really have a deep interpret the reasoning relationship between product attributes and consumer features. Causal inference in big data, there are scholars proposed a new concept, which could have a well result, but no specific study design. This article focuses on data depth causal analysis, to explore a "Big P Big Data" causal inference Big Data causal inference method, and in particular applied to e-commerce personalized recommendation model.Based on the theory of large data, the causal Bayesian network and regression analysis, combined with object background knowledge and theoretical study, we propose a big data causal inference method, and emphasize that as far as possible to collect comprehensive information of research object to build a Big P big data, causal discovery with combined with the background knowledge of causal Bayesian networks and verified by regression analysis of causal inference, the final finish the causal inference of big data. Based on the research background of the research question and the personalized recommendation of electronic commerce, the college students are selected as the research objects. By using the large data causal inference method proposed in this paper, we establish the model of College Students’ online garments recommendation. According to the existing theory of the consumer behavior, building a "big P big data” covering comprehensive information on individual student consumer. Constructing garment styles-brand recommended model and garment dominant color recommended model among college students.This study is further attempts on large data causal inference research, but also on personalized recommendation method further explore. Moreover, there is a certain significance on e-commerce personalized recommendation and big data causal inference.