Survey Visualizer

This repository hosts the source code for a generic visualization tool for data from survey papers, for readers to explore on the browser. It accepts as input a formatted spreadsheet and produces several views in terms of a hierarchy of concepts, document concept similarity, and citation network. See example deployments below.

Setting up locally

Generate files for the Frontend

First, you need to provide a configuration file and the source spreadsheet to generate the data for the frontend.

1. Put all your spreadsheets in this directory and your PDFs (if available) in this directory. Some sample files have been kept there for you. Or use absolute paths (see how to configure in the next step) to point to where your file are.

2. Configure your config.yaml. See here for an example. Detailed instructions on how to configure your system are available here.

3. Then run the following:

user:~$ pip install -r src/compiler/requirements.txt
user:~$ python src/compiler/compile.py --file /path/to/config/file

Bringing up the server

1. Then run the following:
2. Your survey visualizer will show up locally on localhost:3000. 😍
3. For deployment, refer to the active deployments listed below.

user:~$ yarn
user:~$ yarn start

How to Contribute

You can contribute in two forms:

1. Directly to this code base for new features, bug fixes, etc. Open an issue here.
2. To the surveys that pull from this code base, in the form of new paper entries, updates to the taxonomies, and so on. See below for a list of active deployments.

Active Deployments

Topic	Link	Contribute
Virtual, Augmented, and Mixed-Reality for Human-Robot Interaction paper paper	vamhri.com	Contribute
Explainable AI Planning paper paper	explainableplanning.com	Contribute
Model Acquisition for Planning paper	macq.planning.domains	Contribute

How to Cite

download

If you end up using this work, please remember to cite us as follows:

@article{toby,
 title={{TOBY: A Tool for Exploration of Data from Academic Survey Papers}},
 author={Tathagata Chakraborti and Jungkoo Kang and Christian Muise and Sarath Sreedhatan and Michael Walker and Daniel Szafir and Tom Williams},
 journal={arXiv:2306.10051},
 year={2023}}