POTATO: exPlainable infOrmation exTrAcTion framewOrk

Abstract

POTATO is an open-source framework for explainable information extraction from natural language text. It supports the creation, evaluation, and maintenance of rule-based extraction systems, with an emphasis on transparent methods that can be inspected and adapted by users.

TL;DR

• Open-source framework for explainable information extraction.
• Supports rule-based extraction workflows that can be inspected and maintained.
• Early research line behind KR Labs work on transparent NLP systems.
• Available as a paper and open-source repository.

Resources

Cite

@article{kovacs-recski-2022-potato,
  title   = {POTATO: exPlainable infOrmation exTrAcTion framewOrk},
  author  = {Kov{\'a}cs, {\'A}d{\'a}m and Recski, G{\'a}bor},
  journal = {arXiv preprint arXiv:2201.13230},
  year    = {2022},
  url     = {https://arxiv.org/abs/2201.13230}
}