Skip to content

Latest commit

 

History

History
11 lines (6 loc) · 377 Bytes

README.md

File metadata and controls

11 lines (6 loc) · 377 Bytes
Raw

Named Entity Recognition in Brazilian Legal Texts using Conditional Random Fields and Word Embeddings

Running

  • Download LeNER-Br dataset using the script available in lener directory.

    • cd lener && sh download_dataset.sh

  • Run main.py, specifying the NER method and model:

    • python main.py --method CRF --model [FirstOrderCRF|HOCRFAD|HOSemiCRFAD]