mnqu/cotype-pubmed-data-pipeline

Data Pipeline of Lifenet

This is the data pipeline of the Lifenet demo.

To generate the training data and the test data, users first need to download the UMLS database.

Generate Training Data

Users can use the script file "create_training_data.sh" to generate training data. To do that, users may go to the script file, and specify the path of the UMLS database and the output file. Then users may run the script file for data generation.

Generate Test Data

The script file "create_test_data.sh" can be used for test data generation. Specifically, users need to specify the path of the UMLS database, the input corpus and the output file in the script file. Then they can run the script file to generate test data.