Multimodal social network depression detection method based on hypergraph Transform

• Main Authors: LI MINGWEI, LI FENGHUAN, CHEN CHEN, CHEN HAOPENG
• Format: Patent
• Language: Chinese; English
• Published: 08.11.2024
• 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; HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS; PHYSICS; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE,COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTINGPURPOSES, NOT OTHERWISE PROVIDED FOR

The invention belongs to the technical field of natural language processing and graph neural networks, and particularly relates to a hypergraph Transform-based multi-modal social network depression detection method, which comprises the following specific steps of: preprocessing text data and picture data, extracting text features from the text data by utilizing a pre-training language model, and extracting the text features from the picture data by utilizing a pre-training language model; a text hypergraph is constructed in combination with a subject analysis method, and semantic association in text data is fully captured; for image data, feature representation correlation is calculated to construct an image hypergraph, a visual hypergraph convolutional network is introduced, image features are further extracted, and the feature representation capability of the hypergraph is enhanced. According to the method, a hypergraph model capable of reflecting the complexity of multi-modal data is constructed according