Fine-grained Entity Recognition (FgER)

This repository contains the code and data to reproduce the experiments of the paper "Fine-grained Entity Recognition with Reduced False Negatives and Large Type Coverage".

There are three sections of the repository:

1. Datasets (Wiki-FbF, Wiki-FbT and 1k-WFB-g) along with the preprocessed Wikipedia and Freebase.
2. The code used to generate the dataset, i.e., the HAnDS framework code.
3. The code related to the learning models.

Please refer to the individual sections for more details.

Please use the following BibTex code for citing this work.

@InProceedings{abhisheketalAKBC,
  title={Fine-grained Entity Recognition with Reduced False Negatives and Large Type Coverage},
  author={Abhishek, Abhishek and Taneja, Sanya Bathla and Malik, Garima and Anand, Ashish and Awekar, Amit},
  booktitle = {Proceedings of the 1st Conference of the Automated Knowledge Base Construction},
  month     = {May},
  year      = {2019},
  address   = {Amherst, USA},
  publisher = {Automated Knowledge Base Construction},
  url       = {https://openreview.net/forum?id=HylHE-9p6m}
}