Online shopping method and device based on big data and medium

The invention discloses an online shopping method and device based on big data and a medium, and the method comprises the steps: obtaining historical online shopping data in an e-commerce platform, and building a mapping relation between the historical online shopping data and corresponding user information; classifying the historical online shopping data; for each piece of classified historical online shopping data, performing feature extraction according to user information having a mapping relationship with the historical online shopping data to obtain corresponding user identity features, and performing feature extraction according to the user identity features and the historical online shopping data to obtain user behavior features; and performing commodity recommendation for the target user in the online shopping process according to the user behavior characteristics. According to the method, the user preference can be more comprehensively judged under multiple dimensions on the basis of the multiple ca