False news detection system and method based on cooperation of decision tree and common attention

Abstract

The invention discloses a false news detection system and method based on cooperation of a decision tree and common attention. From a new perspective, the invention provides a traditional transparentmachine learning and neural network model combined method, and provides an interpretable false news detection method based on cooperation of a decision tree and common attention, so as to transparently capture powerful fine-grained evidences and explore false parts of false news through the evidences. According to the invention, the false news detection performance is improved, and the transparency of the detection process and the interpretability of the detection result are also provided. According to the method, a transparent and interpretable scheme is provided for a false news detection task, the decision tree model is combined into a common attention network, evidence capable of interpreting claim settlement verification can be provided, and meanwhile the evidence forming process is interpreted through judgment conditions. The system has detachability, two modules of the system can be used for decoupling training. The system has model generalization ability and task staged training ability.

Description

【Technical field】 [0001] The invention belongs to the field of electronic information technology, and relates to a false news detection system and method based on decision tree and joint attention cooperation. 【Background technique】 [0002] The research on fake news detection has roughly gone through two research stages: the first stage is to build a suitable deep model to mine semantic features, emotional features, writing style features, stance features, etc. around the text content of posts or news. social context to extract source-based, post-based, user-based, and network-based credibility indicative features to improve fake news detection performance. Although these methods demonstrate strong effectiveness, they struggle to explain why the detected news is real or fake. To overcome these drawbacks, a popular trend in recent research (the second phase) is to explore evidence-based solutions, which focus on interpretability by capturing relevant pieces of evidence from...

AI Technical Summary

Problems solved by technology

However, although this method shows a certain degree of interpretability, there are still some limitations: first, it is difficult for them to explain the evidence discovery process of news, because this method is a neural network model, which belongs to a non-transparent black box model, Lack of interpretability of the method itself; second, the evidence provided is coarse-grained, usually oriented to the entire sequence of news, lacking focus on the true and false parts of the news sequence

Images