Rule Learner™ is a Machine Learning (ML) component of OpenRules initially developed in 2007. It utilizes powerful ML algorithms for the extraction of patterns from large collections of historical data to present them in the form of readable and executable business rules. Over the years it was successfully applied to real-world projects for automatic discovery of business rules from historical data.
Rule Learner is available on the cloud without any downloads. Click on the button “SaaS Rule Learner” on the right to open the cloud version of Rule Learner with an intuitive web interface:
You will work with the graphical interface shown on the image above. You will be able to select different demo data sets and execute Rule Learner applying different ML algorithms without any registration or subscription. New classification rules will be generated with different metrics describing the quality of these rules. You may download the generated rules as an Excel file. From this interface, you also may download 
the input data sets as Excel files and use them to create your own “custom” data sets.
Work with Custom Models (AWS SaaS Subscription is required)
You also can generate classification rules for your own data sets created in an Excel file similar to sample models. First, you need to click on “Create Account” at the top right corner, enter your email and password, confirm the email, and subscribe to the Rule Learner in AWS Marketplace using our existing AWS Marketplace account. If you already have an account, click on “Sign In“. It will enable you to select your own (custom) data sets and run Rule Learner to produce and download your own classification rules.
This managed service allows you to execute Rule Learner without any installation by paying a minimal fee of $0.25 per generation of rules for your datasets.
You will pay this fee only if/when you actually execute Rule Learner against your own data sets, and all charges will come from Amazon at the end of each month. You can cancel your subscription at any time using the standard AWS “Cancel Subscription”.
Implementation Notes
- This cloud-based version of Rule Learner is implemented as an AWS Lambda Function and, as for all lambdas, the very first execution of the Rule Learner takes a bit more time to wake it up. All consecutive executions are very fast.
- When you provide Rule Learner with your custom Excel file, it will be initially uploaded for processing but after your session is over it will be automatically deleted, so you should not worry about its safety.
- Rule Learner utilizes Machine Learning algorithms provided by the popular open source machine learning framework “WEKA“.
Decision Intelligence Platform 