OASIcs.ICPEC.2024.16.pdf
- Filesize: 0.58 MB
- 12 pages
The use of third-party components (TPCs) and open-source software (OSS) has become increasingly popular in software development, and this trend has also increased the chance of detecting security vulnerabilities. Understanding practical recurring vulnerabilities that occur in real-world applications (TPCs and OSS) is a very important step to educate not only aspiring software developers, but also seasoned ones. To achieve this goal, we analyze publicly available OSS software on GitHub to identify the most common security vulnerabilities and their frequency of occurrence between 2009 and 2022. Our work looks at programming language and type of vulnerability and also analyses the number of code lines needed to be changed to fix different vulnerabilities. Furthermore, our work contributes to the understanding of real-world and human-made data quality required for training machine learning algorithms by highlighting the importance of homogeneous and complete data. We provide insights for both developers and researchers seeking to improve cybersecurity in software education and mitigate risks associated with OSS and TPCs. Finally, our analysis contributes to software education by shedding light on common sources of poor code quality and the effort required to fix different vulnerabilities.
Feedback for Dagstuhl Publishing