Implementation of a simple binary classifier that can determine whether a token is part of a person's name using various techniques for feature extraction, feature indexing, optimization, etc. The data used in this project is derived from the CoNLL 2003 Shared Task on Named Entity Recognition.
To run the project python3 classifier_main.py --model CLASSIFIER
Feature length train : (204567, 136547) Feature length dev : (51578, 136547)
===Train accuracy=== F1: 0.987607
===Dev accuracy=== F1: 0.907864