1. Navigate to the root directory of the project - cd CS399-NLP/
2. Create a virtual environment 2. pip install virtualenv 2. virtualenv venv
3. Activate the virtual environment - source venv/bin/activate
4. Install the requirements - pip install -r requirements.txt

This project is to collect datasets for the task of generating descriptions for charts and graphs. The datasets are collected from the following sources:

1. Wikipedia (Concadia)
2. Statista
3. Pew
4. Accessbility Journals (HCI)

1. Concadia
   1. Download wiki_split.json and resized.zip.
   2. Run concadia.py
   3. Filter manually via visual inspection
2. HCI
   1. Download images.zip and hci.jsonl
   2. Run hci.py
   3. Filter manually via visual inspection
3. Conceptual Captions
   1. Download Train_GCC-training.tsv file
   2. Run conceptual_captions.py
   3. Filter manually via visual inspection
4. Chart-to-Text - Download chart-to-text-dataset.zip from https://github.com/vis-nlp/Chart-to-text
   1. Statista
   2. Pew