This is my submission for third Assignment of the course E1-246: Natural Langauge Understanding. In this assingment we have worked with structural prediction algorithms for the task of Named Entity Recognition on Medical Data.
Two different kinds of models are compared. One based on Conditional Random Fields and the other based on Bi-GRU Transducer Models.
This repository cotains 4 files. 1.) CRF_NER.py : Code file for Conditional Random Fields Model 2.) RNN_NER.py : Code file for Bi-GRU transducer Model 3.) NLU_Assignment3_Report.pdf : A report on the Assignment. 4.) ner.txt : Input file on which both the models are trained
Note: To run the code make sure to provide the correct path of the training file and that all the dependencies/libraries are installed.