In the paper, Visual analytics of set data for knowledge discovery and member selection support, we introduce the general method to build visual analytics system of set data. This repository is to share our implementation of the prototype system build by proposed method in the paper and provide the demo of interactive visualization.
More details about our proposed method are in
- Published version in Decision Support Systems
- Accepted manuscript in arxiv
- How to use this demo after running in your environment (youtube)
Please build the Python environment satisfies the requirements in the pyproject.toml.
If you use poetry, run poetry install.
This demonstration provides a GUI Interface that facilitates knowledge discovery and member selection by clicking topological maps, as shown in the image above. We use the data of 1228 games held in 2018–2019, obtained from Basketball Reference website.
By running learn_and_visualize.py, you can try a prototype of the proposed method's interactive visualization.
learn_and_visualize.py both trains and visualizes the model, but since the trained model is saved in dumped, the visualization is done immediately.
If you want to apply the model to the desired other dataset, code a script to learn model by referring to learn_and_visualize.py.
Please watch this video as a reference
If you have any questions or requests, please give us an Issue!