The invention discloses a character attribute extraction method based on syntactic dependency. The method comprises the steps of S1, text preprocessing; S2, extracting attributes; S3, obtaining a syntactic dependency analysis result of each sentence from the extracted character and character attributes through a Spanford core NLP; S4, for the syntactic dependency analysis result of each sentence,finding the shortest path between the character and the character attribute by marking the character and the character attribute, and storing the vocabularies on the shortest path to obtain a shortestpath vocabulary dictionary; storing the subjects and the vocabularies of the word window sizes before and after the attributes to obtain a word window vocabulary dictionary; S5, respectively calculating the attribute related word dictionary and the Jaccard similarity of the shortest path vocabulary dictionary and the word window vocabulary dictionary to obtain whether the character and the character attribute have a subordination relation or not. According to the method, the character attribute information is extracted from the unstructured English text, a pre-training model is not needed, and the extracted task attribute F1 coefficient is remarkably improved.