This repo contains jupyter notebook implementation of research problem extraction system on NLPContributionGraph (NCG) challenge dataset (training dataset).
Given below is the evaluation results of the implemented BLSTM model on test dataset:
| Metrics (weighted avg) | Precision | Recall | F1_score |
|---|---|---|---|
| Score | 0.81 | 0.78 | 0.79 |
For information regarding the implementation, please look into the notebook :).