Data traceability method and system based on word vector and machine learning

Abstract

The invention provides a data traceability method and system based on word vector and machine learning. The traffic data is collected through the network, file feature extraction is performed based on the traffic data, and a text feature vector is obtained; For each cluster type, the document to be traced is matched with multiple cluster types to obtain the type of the document to be traced; in the type of the document to be traced, the text similarity calculation is performed through the text feature vector, and the source of the document to be traced is traced. determination. The invention has strong scalability, does not depend on a specific system, does not require database transformation, has no requirements on data format, is stable and reliable in traceability, and has high efficiency; Class, and then trace the source by calculating the cosine value, break away from the dependence on the database, and avoid high system management costs.

Description

technical field [0001] The present invention relates to the technical field of information security, in particular to a data traceability method and system based on word vectors and machine learning. Background technique [0002] In recent years, due to the popularization of the Internet and the increase in the number of Internet users, the amount of data generated by people's network activities has also shown explosive growth. While big data provides assistance and value for the development of all walks of life, it also brings new challenges to information security. Especially in some enterprises and various organizations, how to ensure the security of traffic is a very important issue for a large amount of inflow and outflow data. The data traceability technology is such a technology that traces the ins and outs of data, which is also very helpful for data protection and confidential information flow control of various organizations. [0003] Data traceability is a relat...

AI Technical Summary

Problems solved by technology

For example, the document labeling method needs to modify all databases or operating systems in the system, which requires a huge amount of engineering, and may cause additional storage burden, and once the attacker knows the format of the label information, the label information is also easy to be tampered with; reverse The query method is limited to the data traceability of the database, and it is very difficult to construct a reverse query function, so it is not suitable for data traceability in the enterprise environment; the problem with API Hook is that it will generate a lot of redundant and useless data. It is monitoring at the application layer level, so there is no way to analyze the content of the file, resulting in the inability to completely and reliably restore the propagation path and modification records of a specific file

Images