Augmenting Visualizations with Interactive Data Facts to Facilitate Interpretation and Communication

Run a web server:

python -m http.server 8000

Go to http://localhost:8000/ to run the system (preloaded with the cars dataset).

1. Add your dataset in a csv format in the folders dataGenerator/csvs and dataFiles/csvs (already has sample datasets)

2. Create a json file specifying the metadata of your dataset in dataGenerator/dataTypeMaps (already has files for sample datasets). Make sure you specify the "type" field for each attribute and the "isItemAttr" field for the label attribute (e.g. Car Name).

3. To generate the data facts and mappings between facts, visualizations, and annotations, within the dataGenerator folder, execute python mainDataFileGenerator.py csvs/fileName.csv dataTypeMaps/fileName.json. This will create two json files under the dataFiles foder (filename-mainDataMap.json with all the facts and visualizations, and fileName-metadataMap.json which is a modified version of the dataTypeMap file passed earlier).

4. In js/src/main.js, update the paths to the data files passed to the variables dataFileToUse, mainDataMapFileUrl, and metadataMapFileUrl to point to the required csv file and the files generated in Step 4.

5. Go to the root folder, and run a local server (python -m http.server 8000) and go to http://localhost:8000/

@article{srinivasan2019augmenting,
  author={Srinivasan, Arjun and Drucker, Steven M. and Endert, Alex and Stasko, John},
  journal={IEEE Transactions on Visualization and Computer Graphics}, 
  title={Augmenting Visualizations with Interactive Data Facts to Facilitate Interpretation and Communication}, 
  year={2019},
  volume={25},
  number={1},
  pages={672-681},
  keywords={Data visualization;Tools;Visualization;Natural languages;Data mining;Histograms;Complexity theory;Natural Language Generation;Mixed-initiative Interaction;Visualization Recommendation;Data-driven Communication},
  doi={10.1109/TVCG.2018.2865145}
}