Big Data Analysis Used By DARPA Project to Fix Software Security Flaws
The academic researchers are now using bigger data analysis methods to uncover the security flaws in the software. With this analysis, the software flaws are not just found out, but are also fixed.
A nonprofit research lab at the Stanford University, Draper Laboratory, is building a new machine learning arrangement that will carry out detailed checks on the millions of bytes of software code to find flaws in the software security and to fix these flaws. The system is being built by the lab in collaboration by a student group at Stanford University that is led by Andrew Ng. The system has been christened as DeepCode and it has been used before to find out security weakness like Heart bled Bug in OpenSSL. The institute is now expanding the scope of the machine and increasing the amount of data in which Deep Code machine can make its decisions.
The DeepCode system is a new approach to reducing cyber security flaws. It works to collect and consumes huge amounts of software, carries out thorough software search and will index the known bugs as well as security vulnerabilities. It will also identify if the flaws identified are matching to the ones that were previously identified. Draper says that DeepCode must be able to find out what good security codes and bad security codes are as it uses pattern analysis and machine learning techniques to identify the flaws in the security codes.
The main goal of Deep Code system will be to find all the known software bugs. This project is financed by DARPA, (Defense Advanced Research Projects Agency) and the United States Air Force Research Laboratory as part of the MUSE program. The system will be used for carrying out more analysis after the initial platform features are rolled out in the coming months.