E-Commerce Platform Model Based on Heterogeneous Hybrid Model and Maximum Likelihood Estimate

B2C (Business to Customer) is a kind of e-commerce model. The website sells products and services directly to consumers. Under the tide of Internet development, online shopping is popular and sought after by the public, and the market share of B2C online shopping has increased significantly. The number of online goods and services is getting larger and larger. Consumers will face the problem of information overload when shopping on e-commerce platforms. This is accompanied by problems such as difficult selection, longer selection time, and reduced shopping experience. Willingness is not strong. Studying consumer shopping behavior has a positive effect on increasing platform revenue. The consumer's situation will affect their consumption behavior, and mining the consumer's preference and behavior information is helpful to analyze the user's needs. In view of the problems of the above traditional recommendation methods, this paper proposes a recommendation method based on the differentiation of consumer situations, establishes a heterogeneous hybrid model to analyze consumer sensitive situations, and uses the maximum likelihood method to estimate model parameters, provide consumers with references for purchases.