Research on Visual Performance Evaluation Model of E-commerce Websites

Abstract

With the growth of users' personalized demand for e-commerce design and the rapid development of machine learning technology, a large number of Internet pages are automatically designed and generated by computers. After being generated, effective mechanisms are needed for verification and feedback. At present, the of design generally takes traditional web pages as the research object, and there is little research on the complexity and marketing characteristics of e-commerce information. It is necessary to consider how design can promote the effectiveness of communication between users and web pages from the perspectives of psychology, behavior and content. This paper proposes a performance model for the after the generation of contemporary e-commerce web page design. This paper first constructs the marketing system and discusses the circulation mechanism of user demand import, e-commerce web page design, multi-dimensional performance and precision marketing. Under the system, a relational of how visual performance indicators (aesthetic performance, search behavior performance, information usefulness performance) based on scene factors affect overall evaluation is proposed, defined as a multidimensional performance model, and verified by quantitative experiments. The provides design basis for designers and recyclable data for machine learning generated pages.