Multi-feature fusion visual dialogue sentiment analysis method of hybrid model architecture

The invention relates to the technical field of natural language processing, and provides a multi-feature fusion visual dialogue sentiment analysis method of a hybrid model architecture, which comprises the following steps of: acquiring dialogue data containing text information and video information, intercepting the text information according to a statement length, collecting a face image sequence in the video information, and performing sentiment analysis on the face image sequence; obtaining preprocessed text data and image data; based on paired and grouped texts in the preprocessed text data, extracting text features of the preprocessed text data; based on the face image sequence, extracting image features of the preprocessed image data; the text features and the image features are fused, emotion classification is carried out, and emotion categories are obtained; training the sentiment analysis model to obtain a trained sentiment analysis model; according to the method, the effectiveness of conversation e