This repository contains notebooks demonstrating how to deal with the following NLP tasks:
- Named Entity Recognition (NER)
- (cuML) Building SVM with dense word embeddings accelerated by GPU
- (Keras) Building LSTM with dense word embeddings and concatenation with categorical features
- (sklearn) Result analysis
- Negation Scope Detection (NEG)
- (huggingface) Create huggingface.Dataset and evaluation.metric
- (transformers) Fine-tuning BERT for negation scope detection
- (spaCy) Traditional feature engineering approaches
- Visualization for error analysis
- Semantic Role Labeling (SRL)
- (transformers) Two different models on fine-tuning BERT for SRL
See details in the notebooks.