Course recommendation method, device and equipment based on recurrent neural network

Abstract

The invention discloses a curriculum recommendation method, device and equipment based on a recurrent neural network, relates to the technical field of artificial intelligence, and can solve the technical problems of single curriculum recommendation dimension and low accuracy. The method comprises the steps of obtaining first historical course data and object attribute data of a target object; screening an associated object matched with the target object based on the object attribute data, and obtaining second historical course data of the associated object; performing feature conversion processing on the first historical course data and the second historical course data to obtain a first sequence feature of the target object and a second sequence feature of the associated object; the first sequence feature and / or the second sequence feature are / is input into a trained recurrent neural network model, a course prediction result of the target object in a future preset time period is obtained, and the course prediction result comprises a prediction course and a prediction score corresponding to the prediction course; and generating course recommendation information for the target object according to the course prediction result.

Description

technical field [0001] The present application relates to the technical field of artificial intelligence, in particular to a method, device and equipment for course recommendation based on a recurrent neural network. Background technique [0002] Course recommendation is a hot business in the industry, and how to accurately deliver course recommendations has always been a hot research issue that deserves attention. Accurate course recommendation can not only greatly improve the efficiency of manual services, increase user stickiness, but also promote course click transactions. [0003] In the previous course recommendation, user behavior data was mostly collected and processed into a single sample feature, and the sequence information of user behavior was often ignored. As time goes by, the object's earlier behavior information is often covered by new information, which is often ignored in the model, reflecting the characteristics of "time forgetting". As a result, it is i...

AI Technical Summary

Problems solved by technology

[0004] In view of this, this application provides a course recommendation method, device and equipment based on a recurrent neural network, which can solve the problem of being unable to analyze the changing trend of object preferences over time when generating course recommendations, and the analysis dimension is single. Technical problems that lead to inaccurate course recommendations and do not meet the actual course needs of the subjects

Images