An e-commerce recommendation method and system

Abstract

The invention belongs to the technical field of the Internet, and discloses an e-commerce recommendation method. The method includes: obtaining knowledge data and constructing a time-series knowledge map containing multiple different topics, and establishing multi-level labels; labeling according to the latest behavior data of users Extract and construct user dynamic portrait data; associate and match the time-series knowledge map with user dynamic portrait data to identify the map theme that matches the user's current status; analyze the time range corresponding to the user's time-series knowledge map, and compare the corresponding time Products in the range are recommended to users. With the method and system of the present invention, the recommendation results are more accurate and interpretable, and when there is no latest user behavior data, dynamic portraits and product recommendations can still be made to users in the time dimension.

Description

technical field [0001] The invention belongs to the technical field of the Internet, and in particular relates to an e-commerce recommendation method and system. Background technique [0002] At present, the personalized recommendation of major e-commerce websites such as Amazon, JD.com and Taobao is an important way to improve e-commerce brand value, product sales and customer experience. [0003] The mainstream personalized recommendation algorithms include: content-based recommendation, collaborative filtering-based recommendation, and deep learning-based recommendation. However, the method based on collaborative filtering (Collaborative Filtering) only considers the user's feedback data on the product, and the method based on the content (Content Base) only considers the characteristic data of the product, so it is difficult to satisfy the user's deeper level by simply relying on partial data for recommendation. At the same time, there are problems such as sparsity; the...

Images