Code Smells Enabled by Artificial Intelligence: A Systematic Mapping

Abstract

Code smells are an indicator of poor design in software systems. Artificial intelligence techniques have been applied in several ways to improve soft-ware quality in code smells detection i.e. (detection rules or standards using a combination of object-oriented metrics and Bayesian inference graphs). Literature in the field has identified artificial intelligence techniques and compare different artificial intelligence algorithms, which are used in the detection of code smells. However, to the best of our knowledge, there is not a systematic literature review devoted to study in deep the interaction of these fields. In this paper, authors conduct a systematic mapping to get to know how artificial intelligence inter-acts with code smells. Results show the deep connection of Artificial Intelligence with code smells in a solid way, as well as, providing potential challenges and opportunities for future research.