Skip to content

Latest commit

 

History

History
42 lines (24 loc) · 1.27 KB

README.md

File metadata and controls

42 lines (24 loc) · 1.27 KB

Coreference Resolution (In Hindi)

Coreference resolution is the task of linking different mentions of same entity in text. Here we explore rule based and statistical methods for coreference resolution in Hindi.

Getting Started

Prerequisites

You need to have python 2.7 installed on your system.

Installing

  • Run git clone https://github.com/akshatcx/coreference_resolution_cl to clone this repository.
  • Run cd coreference_resolution_cl to enter the project directory
  • Run virtualenv -p python=/usr/bin/python2.7 env to create a virtual environment for the project.
  • Run source env/bin/activate to activate the environment
  • Run pip install -r requirements.txt to install all the dependencies.

Running the tests

  • Run python coref.py to generate the coreference output

Output

We will be getting our data as the output with its respective analysis. It shows the individual pronouns and the word their possible referents.

Authors

  • Akshat Gahoi - IIIT Hyderabad
  • Akshat Chhajer - IIIT Hyderabad

Acknowledgments

  • Thanks to Dr. Dipti Misra Sharma (LTRC IIIT Hyderabad)
  • Mentored by Anirudh Dahiya (LTRC IIIT Hyderabad)
  • Inspired from paper "A Hybrid Approach for Anaphora Resolution in Hindi"(https://www.aclweb.org/anthology/I13-1130)