Rumor detection method combining self-attention mechanism and generative adversarial network

Abstract

The invention discloses a rumor detection method combining a self-attention mechanism and a generative adversarial network. The rumor detection method comprises the steps of collecting rumor text datato form a rumor data set; based on a self-attention mechanism, constructing a generative adversarial network generator comprising a self-attention layer; constructing a discriminator network, and respectively carrying out rumor detection and classification on the original rumor text and the text decoded by the generator; training the generative adversarial network, and adjusting model parametersof a generator and model parameters of a discriminator; and extracting a discriminator network of the generative adversarial network, and performing rumor detection on the to-be-detected text. Compared with an existing rumor detection method, the rumor detection method is higher in detection precision and better in robustness; a self-attention layer is adopted in the generator, key features are constructed through semantic learning of rumor samples, text examples rich in expression features are generated to simulate information loss and confusion in the rumor propagation process, and the semantic feature recognition capacity of the discriminator is enhanced through adversarial training.

Description

technical field [0001] The invention belongs to the field of text recognition, and in particular relates to a rumor detection method combining a self-attention mechanism and a generated confrontation network. Background technique [0002] Rumors have timing and feature diversity, and information will be continuously processed during the dissemination process, which is very confusing. [0003] Rumor detection algorithms based on traditional machine learning need to pre-define and extract text-related features, and are highly dependent on specific types of data sets. Therefore, such algorithms have poor generalization ability. Use neural network structures such as LSTM, GRU, and CNN to extract rumor features to avoid artificial feature engineering. However, this type of model does not combine the characteristics of rumors spreading in the network. In terms of extracting semantic features, the model cannot distinguish key features. Rumors in life will change part of the edge i...

AI Technical Summary

Problems solved by technology

[0003] Rumor detection algorithms based on traditional machine learning need to pre-define and extract text-related features, and are highly dependent on specific types of data sets. Therefore, such algorithms have poor generalization ability

Use neural network structures such as LSTM, GRU, and CNN to extract rumor features to avoid artificial feature engineering. However, this type of model does not combine the characteristics of rumors spreading in the network. In terms of extracting semantic features, the model cannot distinguish key features. Rumors in life will change part of the edge information along with the timing, increasing semantic confusion, thereby reducing the ability of the model to extract key features, resulting in unsatisfactory detection results

[0004] To sum up, the traditional rumor detection algorithm has the problem of unsatisfactory extraction of text semantics and key features, while the general sequence model cannot solve the feature extraction under specific semantics in text detection, resulting in poor generalization ability of the model

Images