Time-series-aware dynamic social scene recommendation method

Abstract

The invention discloses a time-series-aware dynamic social scene recommendation method, which includes: obtaining basic data for analysis from user historical consumption behavior and social behavior records; modeling user time-series consumption behavior and time-series social behavior according to the basic data, thereby Use the obtained dynamic user preferences and dynamic social background information to restore the user's decision-making process in historical consumption behavior, and then combine the user's decision function to estimate the relative ranking of products, and realize the training of relevant parameters in the user's decision function; for new products, Based on the decision function of the completed parameter training, the user's preference score for each product is calculated, and then the result of the user's choice is predicted and recommended through a stable matching method. The above method can not only achieve accurate portraits of users, but also improve the performance of user decision analysis and product recommendation, achieving multiple effects.

Description

technical field [0001] The present invention relates to the field of deep learning and recommendation systems, in particular to a timing-aware dynamic social scene recommendation method. Background technique [0002] The recommendation system is an information filtering system, which aims to analyze user preferences and screen information through user behavior data on e-commerce platforms, so as to provide users with personalized recommendation services. At present, recommendation systems have been widely used in various industries, and the objects that can be recommended include a variety of rich goods and services such as movies, books, music, and news. In recent years, with the development of social platforms and the combination of social elements and emerging business applications, the social behavior between users has become an important basis for recommending products, which means that users' choices on the platform are affected by their social relationships. Therefor...

AI Technical Summary

Problems solved by technology

[0003] In the current research work and patents, no research has been found that combines dynamic social influence and user sequence behavior to recommend users, especially research that incorporates dynamic social influence into the category of time series for representation modeling

Images