Skip to content

README.md

Latest commit

 

History

History
16 lines (10 loc) · 1.36 KB

README.md

File metadata and controls

16 lines (10 loc) · 1.36 KB

Feature Extraction

Feature extraction is the process of root cause analysis using the data set generated by data augmentation. The process consists of two steps. First, the analyzer runs the target program with each input in the dataset and records the state of the program. Second, it compares the traces of the crashed and non-crashed inputs and statistically infers the root cause from the difference.

List of currently available methods

Currently, RCABench supports two types of feature extraction methods as listed in the table below. If you want to add a new method, follow the instructions in "How to add a new feature extraction method".

Method Description
AuroraFE AuroraFE is a method proposed in Aurora [USENIX Security'20].
VulnLocFE VulnLocFE is a method proposed in VulnLoc [ASIA CCS'21].

How to add a new feature extraction method

Basically, new methods can be added in the same way as data extensions, see How to add a new data augmentation method. Also, check AuroraFE and VulnLocFE implementation for more detailed information.