An empirical study on the behavior of e-commerce strategic planning based on deep learning algorithm

On the basis of large-scale literature research, the evaluation and element model for the successful implementation of e-commerce are established, and the key elements (customer, strategy, leadership, technology) and the evaluation elements (system quality, system quality, information quality, service quality) affect the success of e-commerce. First, learn the effective features of the items from the content data through deep learning in advance, and then transform the learned features into the learning task of the collaborative filtering target, and add balance and no relevant constraints to the e-commerce strategic planning behavior values of users and items, using alternating optimization algorithms to learn the value of e-commerce strategic planning behavior and fine-tuning the deep network, and finally get the compact and informative e-commerce strategic planning behavior value of users and items, effectively solving the data sparse problem and cold start in the collaborative filtering algorithm problem. Secondly, the combination of conceptual model and structural equation model has innovated research methods and introduced structural equation model method, which effectively handles the complex relationship between multi-dimensional variables and revises and verifies the hypothetical model. Through path analysis, the interaction and influence between key success factors, success evaluation factors, and successful implementation of e-commerce are explored, and useful attempts are made to expand relevant research data analysis methods.