Development of e-commerce Big data model based on machine learning and user recommendation algorithm

At this stage, the total data volume of e-commerce platforms is extremely large, the overall dimension of data is also particularly high, and the density of value is very low. Traditional prediction and calculation methods are difficult to obtain accurate relevant data, and the machine's own learning ability is very powerful, which can mine some user habits and patterns when using the platform, and then make corresponding plans for future improvement trends. This article will use e-commerce platform user data for experiments, and use the machine's own learning methods and user habits to explore the development of related e-commerce platform Big data. In response to some of the problems existing in existing computing models, the computing model in this article first calculates the user's usage habits during the use process, evaluates the corresponding users through the dissemination power of related items, and constructs a matrix calculation model for user ratings. Then, based on the user's habitual usage behavior, a time interval calculation matrix is established, Finally, the corresponding computing model and decoder are used to calculate the basic information of the user, in order to obtain a more comprehensive information of the user himself. And apply the calculation method proposed in this article to real life, and use this calculation model to build the corresponding e-commerce platform Big data system.