Blockchain illegal information perception method based on deep learning

Abstract

The invention discloses a blockchain illegal information perception method based on deep learning. The method comprises the following steps: 1) collecting information on a chain; 2) extracting text features; 3) extracting picture features; (4) performing feature fusion; (5) using a training set to train a deep learning neural network model in the steps (1) to (4) by adopting a back-propagation algorithm; (6) using the trained neural network model to screen information sorted from the blockchain to screen out illegal information and illegal information. The method is safer, more accurate, and more reliable in judgment.

Description

technical field [0001] The present invention relates to information perception technology of deep learning, in particular to a deep learning-based method for sensing illegal information in blockchain. Background technique [0002] The rapid development and popularization of the network generates massive data, followed by problems such as complex data types and huge information scale. In the field of information perception and identification, the scattered and wide-ranging data generated by illegal organizations and illegal activities, coupled with the gradual increase in unstructured data, have hindered governance actions, which is the focus of attention of security supervision and action departments. [0003] In the field of network illegal information analysis, for the problem of large and fast network information flow, there is currently no particularly efficient solution that can accurately capture and analyze useful information, which mainly involves two factors: first,...

AI Technical Summary

Problems solved by technology

[0003] In the field of network illegal information analysis, for the problem of large and fast network information flow, there is currently no particularly efficient solution that can accurately capture and analyze useful information, which mainly involves two factors: one is the large amount of data, namely It is difficult to determine and obtain the processing object; secondly, due to the rapid development of artificial intelligence technology in recent years, netizens have more contact with traditional models and have been able to avoid the detection of artificial intelligence models by replacing synonyms (homophones, homographs)

Images