Intelligent personality characteristic evaluation method based on multi-source data fusion

The invention discloses an intelligent personality characteristic evaluation method based on multi-source data fusion. The method comprises the following steps: collecting speech video data of a network user; converting the speech video data into a speech text; carrying out user personality characteristic labeling; performing word frequency statistics and sorting on the obtained speech text; for each personality characteristic, enabling psychological professionals to screen out keywords which can best represent the personality characteristic to form a keyword table; processing the speech textand forming word embedding vector features; and inputting the word embedding vector characteristics into a multi-source data fusion model for fusion, and finally outputting a personality characteristic prediction result. Due to the fact that priori knowledge of professionals is combined, the intelligent personality characteristic evaluation method has good interpretability, a more accurate prediction result can be obtaine