Tools for training Stanford NLP's NER models
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
Makefile
README.md
train.prop Add training properties. Aug 23, 2017

README.md

Stanford Named Entity Recognition

Tools for training a Stanford NER model, based on the example in the Stanford NER FAQ.

Setup

  1. Install Java.
  2. Download Stanford NER from https://nlp.stanford.edu/software/CRF-NER.shtml#Download
  3. Extract it to a directory named stanford-ner in this directory.

Usage

  1. Provide two sets of examples in train.txt and test.txt with one example on each line.
  2. Run make sources to produce train.tsv and test.tsv.
  3. Edit train.tsv and test.tsv, replacing O with entity names where appropriate. This may take some time.
  4. Run make model to produce the model, in ner-model.ser.gz.
  5. Run make check to test the model against the examples from test.txt.