Group demand prediction method based on long-term and short-term interests and social influence

Abstract

The invention discloses a group demand prediction method based on long-term and short-term interests and social influence. The method comprises the following steps: firstly, dividing groups based on IP addresses when users access resources; secondly, establishing an interest-score association table for each user in combination with historical access records of the users, training an individual LSTM network model, performing score prediction on a specified item under the condition of comprehensively considering historical behaviors and interest migration of the users, and further constructing a group user-item score table; then, on the basis of analyzing the personality and the professional degree of the users, investigating the sensitive degree of the users to the opinions of other users in the group, and analyzing the mutual relation between the users by mining the intimacy degree between the users in the group, so that a group user model based on social influence is formed; and finally, predicting the group demand according to the model. Long-term and short-term interests of the users and social influence among the users are applied to the demand prediction method, and demands of group members are met to the greatest extent.

Description

technical field [0001] The invention belongs to the technical field of group recommendation, and specifically relates to a group demand prediction method based on long-term and short-term interests and social influence. Background technique [0002] There are still many bottlenecks in public digital cultural services, such as: the lack of unified standards and specifications for resource collection and exchange, the widespread phenomenon of "islanding" of digital resources, the single participation of social subjects in resource construction, and the lack of personalized services. A research breakthrough has been made to address the above-mentioned problems. This invention is a group demand prediction method based on long-term and short-term interests and social influence. Focusing on the improvement of the level of public digital cultural services in our country by focusing on the intelligentization of public digital cultural services for the benefit of the people, it is th...

AI Technical Summary

Problems solved by technology

Moreover, in most existing group demand forecasting methods, user preference is mainly considered, and some other factors that affect group decision-making are less considered, such as individual user interest migration and social influence among group members.

Images