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:
- Datasets (Wiki-FbF, Wiki-FbT and 1k-WFB-g) along with the preprocessed Wikipedia and Freebase.
- The code used to generate the dataset, i.e., the HAnDS framework code.
- The code related to the learning models.
Please refer to the individual sections for more details.