Association Rules Visualization

Goals

• Compute association rules from provided data
• Visualize those rules in a non-standard way
• Accessible as an online tool

Dataset and preprocessing

• data needs to be in CSV format
• datasets from http://repository.seasr.org/Datasets/UCI/csv/ can be used
• support only for nominal values and custom coded binomial values support; all other values are also handled as nominal

Association Rules Mining

• the Apriori algorithm used for rules mining
• open source Apriori Algorithm implementation in JavaScript used to run in NodeJS environment

Postprocessing

• the Apriori.js open source project only provides confidence for rules and support for item sets
• my custom postprocessing step computed support, lift and conviction for all rules

Technologies

• IBM Bluemix - for cloud NodeJS environment
• Node-RED tool - for quick prototyping and visual data flow scaffolding
• D3.js - for creating visualizations in the browser
• Apriori.js - for extracting rules from a dataset

Solution URLs

• UI - http://association-rules.eu-gb.mybluemix.net/ui/
• Node RED - http://association-rules.eu-gb.mybluemix.net/red/#
• Presentation - https://docs.google.com/presentation/d/1nK3FydLWRTjP_7R4pVElk4-Mz3RkiRV60EyYw7cR-iY/pub?start=false&loop=false

Sources

• https://console.ng.bluemix.net/
• http://nodered.org/
• https://d3js.org/
• https://github.com/seratch/apriori.js
• https://en.wikipedia.org/wiki/Association_rule_learning
• https://en.wikipedia.org/wiki/Apriori_algorithm