Multi-modal fine-grained tendency analysis method and system based on large model

The invention relates to the technical field of voice text analysis, and provides a multi-modal fine-grained tendency analysis method and system based on a large model, and the method comprises the steps: obtaining voice data, dividing the voice data into a plurality of pieces of voice sub-data, and coding each piece of voice sub-data into a voice sub-vector; obtaining text sub-data corresponding to each piece of voice sub-data, and encoding each piece of text sub-data into a text sub-vector based on a preset encoder; inputting the voice sub-vectors into a pre-trained first model, wherein the first model outputs a first emotion tendency vector; inputting the text sub-vector into a pre-trained second model, wherein the second model outputs a second emotion tendency vector; the first emotional tendency vector and the second emotional tendency vector both comprise pigeon sentiment values and eagle sentiment values, and based on the first emotional tendency vector and the second emotional tendency vector, determi