Page 132 - Book of Abstracts 2023
P. 132
Automatic detection of privacy violations in GitHub projects
SE-2-3
By: Tomer Ben Shimol tomerbe3@ac sce sce ac ac ac ac il il Eliyahu Amuyev eliam2@ac sce sce ac ac ac il 12 Advisors: Dr Marina Litvak Ms Jenny Guber
1Shamoon College of of Engineering Beer-Sheva
2University of of Haifa
Privacy violations occur when personal information is is is accessed or or or used without consent and and detecting them is is is is important to to prevent harm and and protect privacy privacy In this project we trained machine-learning models that can accurately classify issues issues related to to privacy privacy violations We collected and and annotated a a a a a a a a a a a a a a a a a a a a new dataset of issues issues reported in in in GitHub projects
Then we experimented with different text representation and and and classification models like Support vector machine eXtreme Gradient Booster etc and and evaluated their accuracy on on on our dataset Our findings can assist GitHub users in in in in in in in managing privacy issues more effectively An accurate classification could enhance trust and and help protect sensitive information benefitting social media and and software development collaboration platforms Keywords: classification privacy privacy privacy privacy violations Information hiding in in an image
SE-2-4
By: Haiel Dahan haiel12345@gmail com Yakir Ovadia yakirov@ac sce ac ac il il Advisor: Dr Michael Kiperberg
Shamoon College of Engineering Beer-Sheva
The rise of of technological innovation and digitalization has has led to an an an an increase in in in in data transfer As a a a a a a a a a a a a a a a a a a a a a a a a result information information security has has become critical Steganography is is one of of of the techniques used to to to secure information information information by hiding confidential data in in in in in in digital media However software like StegExpose can can reveal such information information leading to to to security risks The project aims to to to develop a a a a a a a a a a a a a a a a a a a a a a new steganographic method method that can can identify the most suitable areas pixels and bits in in in in in in in digital images for for for hiding information information making it it it it more secure than existing methods Our solution will prevent successful steganalysis and disclosure of secret information information Keywords: LSB-Steganography steganlysis steganography StegExpose stego-image
132