CRFEntityRecognier uses CRF model to do named entity recognization (NER) in preprocessed texts.
python3
pandas
sklearn
sklearn_crfsuite
scipy
joblib
./CRFEntityRecognier.sh <-i infile>
[-n n_jobs]
[-o outdir]
-i: filename of raw input text. Input text shoud be in TSV format with 4 colomns as follows:
Sent_ID Word Pos Tag
-n: number of threads to use. Default=1
-o: directory to save trained model. Default=./model
Use following command to test:
./CRFEntityRecognier.sh -i testdata/test.tsv