Project on visualizing association rules extracted from covid-19 data.
APACHE-2.0 License
This project involves applying data mining techniques on COVID-19 data to extract association rules and visualize these rules as a network. This repository includes code and supplementary materials for the project and is available online at https://github.com/apetkau/comp7944-project.
Prepared for COMP 7944 at the University of Manitoba on April 23, 2020 by:
Below lists all the interactive versions of our visualization of association rules (as a network). For each dataset we produced two networks, one where nodes are colored by confidence and the other where nodes are colored by lift. Zooming or panning can be accomplished using the mouse, and network nodes can be dragged and dropped.
Copies of the two datasets we are using can be found at:
We processed the above datasets to define sets of items (transactions) for use with the Apriori algorithm for finding frequent itemsets and association rules. We constructed 4 separate transactional itemsets for further processing. Jupyter notebooks for processing this data is given below.
We next applied data mining techniques to find association rules in the above datasets and visualize the rules. Jupyter notebooks for this process are given below.
To reproduce this analysis you can use the following instructions to install dependencies using conda (though we note some additional R packages may need to be installed manually).
Install Miniconda used for software dependency management.
Install dependencies (from dependencies.conda
file) using the command:
conda create --name datamining --file dependencies.conda
Activate the conda environment with installed software:
conda activate datamining
Run Jupyter lab.
jupyter lab
You should now be able to load up the Juptyer notebooks and work with them.
The source data for this project (under the data/
directory) is redistributed under the respective licenses of the original providers. The code in this project is distributed under the Apache 2.0 license.