abstract |
The present invention provides a method for extracting medical entity relationship based on hierarchical reasoning, which is a method for extracting entity relationship triples related to novel coronavirus pneumonia through an entity relationship extraction model based on hierarchical reasoning. The related literatures were aligned to construct the COVID‑19 entity relation extraction dataset; then, based on gated convolution and dilated convolution, a feature extractor was constructed to extract contextual information from different dimensions; finally, based on the hierarchical reasoning mechanism, using The document information of different granularities is used to extract the entity relationship from the text, which improves the accuracy of the document-level entity relationship extraction model. The beneficial effects of the present invention are: extracting the entity relationship of the new coronary pneumonia related data can provide a basis for subsequent research tasks such as knowledge maps, question answering systems, etc., and provide a basis for further prevention and treatment of the new coronary pneumonia. |