ICSME 2025
Sun 7 - Fri 12 September 2025 Auckland, New Zealand

In software development environments, code quality is crucial. This study aims to assist Machine Learning (ML) engineers in enhancing their code by identifying and correcting Data Leakage issues within their models. Data Leakage occurs when information from the test dataset is inadvertently included in the training data when preparing a data science model, resulting in misleading performance evaluations. ML developers must carefully separate their data into training, evaluation, and test sets to avoid introducing Data Leakage into their code. In this paper, we develop a new Visual Studio Code (VS Code) extension, called LEAKAGEDETECTOR, that detects data Leakage — mainly overlap, preprocessing and multi-test leakage – from Jupyter Notebook files. Beyond detection, we included two correction mechanisms named Quick Fix: a conventional approach that manually fixes the leakage and an LLM-driven approach that guides ML developers toward best practices for building ML pipelines. The plugin and its source code are publicly available on GitHub at https://github.com/SE4AIResearch/ DataLeakage Jupyter Notebook Fall2024. The demonstration video can be found on YouTube: https://www.youtube.com/watch?v=uyLLaxutzsg&t=6s. The website can be found at https://leakage-detector.vercel.app/.

Fri 12 Sep

Displayed time zone: Auckland, Wellington change

13:30 - 15:00
Session 16 - Security 2Research Papers Track / Industry Track / Registered Reports / NIER Track at Case Room 2 260-057
Chair(s): Gregorio Robles Universidad Rey Juan Carlos
13:30
15m
Understanding the Faults in Serverless Computing Based Applications: An Empirical Study
Research Papers Track
Changrong Xie National University of Defense Technology, Yang Zhang National University of Defense Technology, China, Xinjun Mao National University of Defense Technology, Kang Yang National University of Defense Technology, Tanghaoran Zhang National University of Defense Technology
13:45
15m
Security Vulnerabilities in Docker Images: A Cross-Tag Study of Application Dependencies
Research Papers Track
Hamid Mohayeji Nasrabadi Eindhoven University of Technology, Eleni Constantinou University of Cyprus, Alexander Serebrenik Eindhoven University of Technology
14:00
15m
Trust and Verify: Formally Verified and Upgradable Trusted Functions
Research Papers Track
Marcus Birgersson KTH Royal Institute of Technology, Cyrille Artho KTH Royal Institute of Technology, Sweden, Musard Balliu KTH Royal Institute of Technology
14:25
10m
MalLoc: Towards Fine-grained Android Malicious Payload Localization via LLMs
NIER Track
Tiezhu Sun University of Luxembourg, Marco Alecci University of Luxembourg, Aleksandr Pilgun University of Luxembourg, Yewei Song University of Luxembourg, Xunzhu Tang University of Luxembourg, Jordan Samhi University of Luxembourg, Luxembourg, Tegawendé F. Bissyandé University of Luxembourg, Jacques Klein University of Luxembourg
Pre-print
14:35
15m
Levels of Binary Equivalence for the Comparison of Binaries from Alternative Builds
Industry Track
Jens Dietrich Victoria University of Wellington, Tim White Victoria University of Wellington, Behnaz Hassanshahi Oracle Labs, Australia, Paddy Krishnan Oracle Labs, Australia
14:50
10m
Repairing vulnerabilities without invisible hands. A differentiated replication study on LLMs
Registered Reports
Maria Camporese University of Trento, Fabio Massacci University of Trento; Vrije Universiteit Amsterdam