// Educational Data Mining: learning Design Rules

Welcome!

The purpose of this blog is to keep track of weekly progress and reflection on the process of researching at George Mason University's REU site on educational data mining. Information about the author can be found under About, research-related content under Posts, and other fun happenings under More.

A little about design rules and how to learn them:

As a newcomer to a software development project with millions of lines of code, learning design rules is a time-consuming and sometimes overwhelming task, especially when code is documented poorly. We can imagine a tool that automatically authors such design rules and presents them to the user, along with examples of code that obey or possibly violate the rule, helping the developer to become more familiarized with the codebase in less time. The greatest challenges to building this tool involve choosing how to properly represent lines of code and quantify their features, finding an algorithm with sufficient scalability to handle millions of lines of code, and above all, understanding what makes a design rule relevant or interesting.