AMASS: Automated Software Mass Customization via Feature Identification and Tailoring
Keywords:Program customization, Deep learning, Binary analysis
The rapid inflation of software features brings inefficiency and vulnerabilities into programs, resulting in an increased attack surface with a higher possibility of exploitation. In this paper, we propose a novel framework for automated software mass customization (AMASS), which automatically identifies program features from binaries, tailors and eliminates the features to create customized program binaries in accordance with user needs, in a fully unsupervised fashion. It enables us to modularize program features and efficiently create customized program binaries at large scale. Evaluation using real-world executables including OpenSSL and LibreOffice demonstrates that AMASS can create a wide range of customized binaries for diverse feature requirements, with an average 92.76% accuracy for feature/function identification and up to 67% reduction of program attack surface.
How to Cite
Copyright (c) 2022 EAI Endorsed Transactions on Security and Safety
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
This is an open-access article distributed under the terms of the Creative Commons Attribution CC BY 3.0 license, which permits unlimited use, distribution, and reproduction in any medium so long as the original work is properly cited.
Office of Naval Research
Grant numbers N00014-15-1- 2210