Long sequence prediction method and device for microblog topic trend and computer storage medium

• Main Authors: DENG AOBO, YUAN LIU, HUANG LINGYUE, JIANG HAO, ZHANG YI, LI HAO, CHEN CHENG
• Format: Patent
• Language: Chinese; English
• Published: 29.03.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FORADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORYOR FORECASTING PURPOSES; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention provides a long sequence prediction method aiming at a microblog topic trend, which comprehensively considers graph structure space characteristics and time characteristics of microblog topics in a propagation evolution process, performs time slicing on a forwarding network in a topic evolution process, takes node characteristics as attribute input of nodes, and performs prediction on the microblog topics. Constructing a dynamic graph sequence of the topic forwarding network; a GAT graph attention network is used to carry out graph structure space information learning representation on a forwarding network in the topic; performing feature fusion on the represented graph structure space information and statistical feature information of topic evolution to form a fused feature sequence; according to the method, the calculation complexity of the model can be effectively reduced, and longer sequence input and output can be realized on the premise of ensuring the performance of the model. Compared wi