Sequence recommendation method based on user behavioral difference modeling

Abstract

The present invention discloses a sequence recommendation method based on user behavioral difference modeling. The method comprises: acquiring historical behavior information of a user; calculating acommodity feature vector according to the acquired historical behavior information; by combining a commodity feature vector, using a behavioral difference modeling method for sequence modeling, and obtaining the current demand and historical preferences of the user by using two different neural network architectures; and according to the current purchase demand and the historical preferences of the user, predicting the next commodity of interest to the user through joint learning, performing matching in a commodity vector space, finding a plurality of commodities that are closest to the predicted result in the commodity vector space, and generating a commodity recommendation sequence. According to the method provided by the present invention, through difference modeling on user timing sequence behaviors, the current demand and long-term preferences in the purchase decision of the user can be intelligently understood, and accurate sequence recommendation services can be provided for users.

Description

technical field [0001] The invention relates to the technical fields of machine learning and e-commerce, in particular to a sequence recommendation method based on user behavior distinction modeling. Background technique [0002] With the continuous development of online shopping platforms, recommender systems have become an irreplaceable and important part of e-commerce. The recommendation system can learn the preference information hidden in the user's historical behavior, so as to further predict the user's shopping behavior, help customers choose satisfactory products, and promote the revenue of e-commerce platforms. Therefore, how to efficiently and accurately provide users with personalized product recommendation services has always been an important research issue in academia and industry. [0003] At present, there are two main categories of research on recommender systems: [0004] 1) Recommendation system based on user static preference [0005] Content-based, c...

AI Technical Summary

Problems solved by technology

This short-session sequence recommendation method can model the dynamic changes of user behavior in the short term through deep neural networks, but this method ignores the user's own preferences, which makes the recommendation results often meet the user's needs but not the user's favorite type.

At the same time, neither method can deeply model the dynamic changes in the entire decision-making process of users, and does not specifically analyze the different degrees of preference expressed by different behaviors of users.

Therefore, it is difficult to accurately model the complete decision-making process when users choose goods or services by using existing recommendation methods, and the user needs and preferences cannot be combined, resulting in recommended content that fails to meet user expectations

Images