Accurate identification method for APP file

Abstract

The invention relates to the technical field of file type identification, in particular to an accurate identification method for an APP file, which comprises the following steps: reading a binary data stream of a to-be-identified target file and extracting a basic feature and a corresponding nominal suffix; determining an implicit file format according to the head sequence to construct a mapping relation library, and determining a suffix matching degree according to a nominal suffix extraction result and the mapping relation library; determining a file processing mode according to the suffix matching degree so as to determine whether the final file type of the to-be-identified target file is determined again or not; according to a result needing to be re-determined, determining semantic features according to the readable character string, generating a node attribute vector according to the semantic features, the head sequence and the structure mark, and obtaining a global attribute vector; associating and integrating the node attribute vector and the global attribute vector to construct heterogeneous graph structure data; and based on the APP classification scoring model, scoring the heterogeneous graph structure data and outputting a confidence coefficient vector, and determining a final file type according to the confidence coefficient vector. According to the invention, the APP file type can be effectively identified.

Images