Key information identification and extraction method and system

The invention relates to a key information identification and extraction method and system, and belongs to the technical field of information processing. The method comprises the steps that firstly, atarget path selected by a user is acquired, a folder structure is analyzed, and a corresponding label tree is generated; each node content of the label tree is traversed, and a storage directory is generated in a database; documents or pictures in a folder selected by a user are identified, the documents or pictures are converted into binary data to serve as to-be-classified samples, and the to-be-classified samples are compared with each sample in a sample data set to obtain sample distances; k samples most similar to the to-be-classified samples are found, and labels of the K samples are obtained; selecting the classification with the highest occurrence frequency in the K sample labels as the classification of the to-be-classified samples; and obtaining the identification content according to the labels corresp