A network learning group division method based on the learning-generated network similarity

Abstract

The invention discloses a network learning group division method based on the similarity of learning generation networks. The method comprises the following steps of 1) establishing a user knowledge point association network, and calculating the time sequence relevancy of an i+1th knowledge point and previous i knowledge points in a user learning sequence; 2) constructing the learning generation networks of users; 3) acquiring the content similarity PLGNCntDis ux.uy between PLGNux and PLGNuy of the learning generation networks; 4) calculating the structural similarity PLGNStrDis ux.uy betweenPLGNux and PLGNuy of the learning generation networks of the users; 5) using a weighted sum result of the content similarity and the structural similarity as the overall similarity of the learning generation networks of the users, clustering the learning generation networks of the users by adopting a CURE hierarchical clustering algorithm based on the similarity according to the overall similarityof the learning generation networks of the users, and achieving network learning group division based on the similarity of the learning generation networks, wherein PLGNux, PLGNuy, PLGNCntDis ux.uy and PLGNStrDis ux.uy are all shown in the descriptions. According to the division method, the network learning group division is achieved by considering the leaning process and cognitive characteristics of the users.

Description

technical field [0001] The invention relates to a group division method of network learning users, in particular to a network learning group division method based on network similarity generated by learning. Background technique [0002] Most recommender systems mainly focus on the recommendation of a single user, however in many daily activities it is necessary to make recommendations for groups formed by multiple users. In recent years, the group recommendation system (Group recommendation system) has gradually become one of the research hotspots in the field of recommendation systems. How to integrate the preferences of group members to meet the preferences of members and divide groups is the main task of group recommendation. [0003] Specifically, group division refers to assigning users to different groups, so that users in each group have the same characteristics in some aspects. Wang Zhongqing proposed a hidden factor graph model, using various implicit and explicit...

AI Technical Summary

Problems solved by technology

However, the existing features often lack the consideration of user cognition and cannot fully describe the entire learning process of users, which affects the accuracy of group division to a certain extent.

Images