Research And Implementation On Product Keyword Optimization Based On User's Behavior

In recent years, with the rapid development of the Internet economy, B2B business platform plays increasingly important roles in the Internet economy, which links manufacturing enterprise with consumer industry. Descriptive keywords released by manufacturing enterprise in the platform directly exert an influence on whether users can retrieve the exact products, which is critical for the operation of manufacturing industry. How to improve the visibility of the keywords released by manufacturing industry directly reflects the service effect of electronic business platform. Therefore, combined with the actual demand of electronic business platform, this thesis studies the optimization of product keywords based on user's behavior, to realize prototype system. The goal of product keywords optimization recommendations aims at promoting the individuation and customization of e-commerce platform service and improving the quality of service.The main works are as follows:(1) Business platform have accumulated large amount of users' access information, combining with mining users' search behavior pattern, to determine the target data source, and pretreatment, a weighted trajectory data set construction method is proposed, both the size of the data and reflection of the idea of users' behavior patterns, generating mining data sets.(2) Combined with the user feature and product searching field to classify users, design frequent item sets mining method to acquire the association between all kinds of users' searching keywords and their attention to product keywords. On the basis of this, designing the clustering algorithm to generate keyword association lexicons and hot lexicons from the angle of semantic relevance and association degree.(3) Based on the constructed association lexicons and hot lexicons, design keywords recommendation method on the basis of weight k neighbor thought, to achieve optimized keywords recommendation for target product. And design evaluation mechanism to evaluate the effect of optimized keywords on the basis of association datasets, to verify the effectiveness of the method.On this basis, the product keywords optimization system based on user's behavior is realized. Help users to develop a reasonable product description keywords and improve the product's visibility. At present, the system has been commissioned by the R & D acceptance, under deployment of applications.