steveneale / ner_crf

Jupyter Notebook describing named entity recognition (NER) using conditional random fields (CRFs), implemented using scikit-learn / sklearn-crfsuite.

ner_crf

ner_crf is a Jupyter Notebook describing named entity recognition (NER) using conditional random fields (CRFs), implemented using scikit-learn/sklearn-crfsuite.

Dependencies

ner_crf is written in Python, and so a recent version of python3 should be downloaded before using it. Downloads for python can be found at https://www.python.org/downloads/ (version 3.5.1 recommended). You also need to be able to run Jupyter Notebooks (see Project Jupyter).

The following python libraries will also be needed to run the ner_crf notebook:

• pandas
• NumPy
• scikit-learn
• sklearn-crfsuite
• matplotlib
• PyPrind