Reproduction package of the paper "DeepCVA: Automated Commit-level Vulnerability Assessment with Deep Multi-task Learning" in Automated Software Engineering (ASE) 2021
Hello, I'm trying to reproduce the result. I found that the distributions of CVSS metrics in the file cvss_map.csv is different from the paper. I would appreciate it if you could provide the dataset in your paper.
Hello, I'm trying to reproduce the result. And while running infer_features_sequential.py, it requires TextProcessor from text_processing. However, there is no such local file. And the pip package text_processing doesn't work. How should I fix this?
Dear Author :
As I check the dataset in data/java_vccs.csv, It seems that labels don't contain Severity?(they only have 6: AV/AC/Au/C/I/A. But it seems don't have Severity) .
Secondly, if the label is not None, (ie. Confidentiality = partial) you will consider this commit may has this kind of Vulnerability?
Is there a binary classification (classify to yes or not) for each type of vulnerability not a triple classification?