Author : Sudip Mahajan, Dr. Geetanjali V. Kale and Sudarshan Bhide

Software Cloning as an inconsistency in source code has attracted lot of attention across Software Engineering research community. Code Cloning may have adverse effects on software development process and hence a developer should be aware about its facets. Present work aims to highlight challenges in processing Abstract Syntax Tree and applying deep learning approaches on AST representation of source code towards establishing semantic proximity of code fragments. The SeSaMe dataset is used in this work for code clone detection. Results are promising and encourage using larger datasets.

Keywords: Software clone; code clone detection; deep learning