This project is based in term recognition using BERT for the NLP domain. To see more of our methodology, results and analysis, see the report.
By Averie Ho Zoen So and Silviya Silwal
Use the requirements.txt file to install all necessary packages.
$ pip install -r requirements.txt
- Use
python bert_pipeline.py
to run experiments for bert. Hyperparameters are pre-defined in the code. - Use
python nltk_spacy_iob.py
to run off-the-shelf baseline experiments. - Use
python max_rule_baseline.py
to run the max rule baseline. - Use
python ner_span_eval.py
to conduct span-based evaluation and error analysis. - Use
python cohens_kappa.py
to measure inter-annotator agreement.