udrasht/neural-dependency-parser-hindi

Neural Dependency Parsing

This repository contains code for a project on Neural Dependency Parsing. The project involves parsing raw data, preprocessing it, and implementing two different models for dependency parsing: Support Vector Machine (SVM) and Feed-forward Neural Network.

Code Structure

• code/: This folder contains the implementation code.
  • data/: Raw data resides here.
  • data_preprocessing/: Contains code for preprocessing the raw data to extract useful information.
  • SVM/: Implementation code for the Support Vector Machine model.
  • Feed-forward neural network/: Implementation code for the Feed-forward Neural Network model.

SVM Implementation Notes

In the SVM code, all file paths are relative. Make sure to adjust them according to your file system.

Feed-forward Neural Network Implementation Notes

In the Feed-forward Neural Network code, the file paths for representing word embeddings are relative. You need to download the pre-trained model from IndicNLP and adjust the paths accordingly.

Requirements

Make sure you have the following dependencies installed:

• Python
• json
• re
• sklearn
• scipy
• torch
• numpy
• matplotlib
• seaborn
• fasttext

Report

For more detailed information about the project, please refer to the report.pdf file located at the root of this repository.

For Complete Project refer to the Repo: link