Deep learning method for recognizing cause-effect relationship of contradictory dispute event

The invention relates to a deep learning method for recognizing the cause-effect relationship of a contradictory dispute event. The method comprises the following steps of 1, obtaining a training corpus text; 2, preprocessing a training corpus text; 3, automatically labeling each extracted sentence array, and outputting M labeled sentence arrays containing one result sentence, one reason sentenceand 15 random sentences; 4, training a causal relationship recognition model based on a context attention mechanism; 5, preprocessing the to-be-recognized conflict event description text, and outputting semantic feature vectors of corresponding sentences; 6, combining the semantic feature vectors, and inputting the combined semantic feature vectors into a trained causal relationship recognition model; and step 7, outputting the causal relationship between the contradictory dispute events. According to the technical scheme, the problem that the causal relationship recognition accuracy of a traditional classification me