Causal event extraction method based on causal event extraction model

The invention discloses a causal event extraction method based on a causal event extraction model, and the method comprises the following steps: constructing a causal event extraction model based on two subtasks: event extraction and causal relationship recognition; obtaining a to-be-input statement; coding a to-be-input statement on the basis of the pre-training language model, decoding the coded to-be-input statement on the basis of a sequence labeling decoder, and extracting an event with causal association; constructing an initial background image based on a causal event extraction model; inserting an event into the initial background image to obtain an updated background image; encoding and updating the updated background image by adopting a graph neural network to obtain the expression of the extracted event; and based on the classifier and the extracted representation of the events, obtaining a causal relationship between the events, and realizing extraction of the causal events.