Random Forest for Code Smell Detection in JavaScript

Abstract

JavaScript has become one of the most widely used programming languages. JavaScript is a dynamic, interpreted, and weakly-typed scripting language especially suited for the development of web applications. While these characteristics allow the language to offer high levels of flexibility, they also can make JavaScript code more challenging to write, maintain and evolve. One of the risks that JavaScript and other programming languages are prone to is the presence of code smells. Code smells result from poor programming choices during source code development that negatively influence source code comprehension and maintainability in the long term. This work reports the result of an approach that uses the Random Forest algorithm to detect a set of 11 code smells based on software metrics extracted from JavaScript source code. It also reports the construction of two datasets, one for code smells that affect functions/methods, and another for code smells related to classes, both containing at least 200 labeled positive instances of each code smell and both extracted from a set of 25 open-source JavaScript projects.