DecisionDroid: a supervised learning-based system to identify cloned Android applications

Abstract

This study presents DecisionDroid, a supervised learning based system to identify cloned Android app pairs. DecisionDroid is trained using a manually verified diverse dataset of 12,000 Android app pairs. On a hundred ten-fold cross validations, DecisionDroid achieved 97.9% precision, 98.3% recall, and 98.4% accuracy.

BibTeX

@inproceedings{Kohli:FSE17,
  author    = {Ayush Kohli},
  title     = {{DecisionDroid:} a supervised learning-based system to identify cloned Android applications},
  booktitle = {{ESEC/SIGSOFT} {FSE}},
  pages     = {1059--1061},
  publisher = {{ACM}},
  year      = {2017},
}

DecisionDroid: a supervised learning-based system to identify cloned Android applications

Abstract

Related papers