Update December 2022

The backend server is no longer hosted - it is easy to host your own backend and frontend server locally.

RepGraph

RepGraph is an application tool to equip Natural Language Processing (NLP) Researchers and Engineers with the ability to visualise and analyse graph data that has been processed by one of the MRP frameworks. Demonstration Video - Demonstration Website

Last Updated March 2021

Frameworks Supported

• DMRS
• EDS
• PTG
• UCCA
• AMR

Built With

• React
• React visx
• Material UI
• Spring Boot
• Java

Contents

• RepGraph
  • Contents
  • Usage
  • Getting started
  • Repository structure
  • Contribution guidelines

Usage

The user interface is easy to use and intuitive but for a clear and easy usage guide, refer to the user manual here: User Manual

If you simply want to use the application without the hassle of compiling and running everything locally - you can go here

For access to sample data as well as a description of Uniform Graph Interchange Format: http://mrp.nlpl.eu/2020/index.php?page=14

Getting Started

To use the RepGraph tool is simple. All that is needed are up to date versions of the following prerequisites and Language compilers/interpreters.

• To access source code, simply clone the RepGraph repository to your local file system:

git clone https://github.com/RepGraph/RepGraph.git

Installation

Prerequisites

• Java
• Maven
• Node Package Manager
• Python 3.8.5

Backend

pip install -r requirements.txt
mvn compile
mvn package
java -jar *Location of Jar File*

Frontend

npm install
npm start

Remember to change the API endpoint in the AppContextProvider file if you want to use a locally hosted backend.

Repository structure

RepGraph
├── src 'contains all source code for the backend and frontend'
|   ├── main 'contains source code for the repgraph tool'
|       ├── java 'contains all java source code that is used in backend'
|       ├── javascript 'contains all javascript code that is used in frontend'
|       └── resources 'contains all other resources used such as config and property files'

Acknowledgements

A big thanks to Jan Buys (jbuys@cs.uct.ac.za) for his help and supervision of this project.

Authors

• Edan Toledo
• Roy Cohen
• Jaron Cohen
• Jan Buys

Contact

• TLDEDA001@myuct.ac.za
• CHNROY002@myuct.ac.za
• CHNJAR003@myuct.ac.za
• jbuys@cs.uct.ac.za

Contribution guidelines

Please ensure that you adhere to the following guidelines when making any contributions to this repo:

• First create an issue describing the feature, improvement or bug fix. Issue wording should be clear and easy to understand.
• Create a new branch for the specific issue and prefix your branch name with the issue number followed by a few words describing the issue (e.g. 34-dark-mode).
• Please also ensure all paths are relative and not hard coded
• For all newly created functions, please make sure to include docs
• For all newly created functions, include corresponding unit tests in the test folder.
• Commit any changes to your branch with an appropriate message.
• Before creating a new merge request, ensure that your commit passes all unit tests.
• Create a new merge request. Your contribution will not be accepted unless:
  1. all new functions have docs,
  2. the appropriate unit tests have been created,
  3. and your code passes all unit tests.
• Once a core member has reviewed and approved with the changes/additions you have made, your branch will be merged with master and will be live in the RepGraph tool.
• Please make sure old branches are deleted after they have been merged and are no longer in use.

Citation

If this work helped you in anyway please consider citing us:

@inproceedings{cohen-etal-2021-repgraph,
    title = "{R}ep{G}raph: Visualising and Analysing Meaning Representation Graphs",
    author = "Cohen, Jaron  and
      Cohen, Roy  and
      Toledo, Edan  and
      Buys, Jan",
    booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
    month = nov,
    year = "2021",
    address = "Online and Punta Cana, Dominican Republic",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.emnlp-demo.10",
    pages = "79--86",
    abstract = "We present RepGraph, an open source visualisation and analysis tool for meaning representation graphs. Graph-based meaning representations provide rich semantic annotations, but visualising them clearly is more challenging than for fully lexicalized representations. Our application provides a seamless, unifying interface with which to visualise, manipulate and analyse semantically parsed graph data represented in a JSON-based serialisation format. RepGraph visualises graphs in multiple formats, with an emphasis on showing the relation between nodes and their corresponding token spans, whilst keeping the representation compact. Additionally, the web-based tool provides NLP researchers with a clear, visually intuitive way of interacting with these graphs, and includes a number of graph analysis features. The tool currently supports the DMRS, EDS, PTG, UCCA, and AMR semantic frameworks. A live demo is available at https://repgraph.vercel.app/.",
}