Multimodal Classification in Social Media: Tensor and Neural Network Feature Fusion

With the rapid development of internet technology, the social media sector has seen an increasing amount of multimodal data, including text, images, and videos, that contain a large amount of useful information. These data play an extremely important role in sentiment analysis, personalized recommendations, and public opinion monitoring. Both in industry and academia, the field of multimodal data has received unprecedented attention. This paper focuses on the problem of multimodal classification of text and images in the social media sector. From the perspectives of tensor feature fusion and neural network feature fusion, two novel models for multimodal classification are proposed.