Labelled Vulnerability Dataset on Android source code (LVDAndro) to develop AI-based code vulnerability detection models. [Dataset]

Senanayake, Janaka; Kalutarage, Harsha; Al-Kadri, Mhd Omar; Piras, Luca; Petrovski, Andrei

Abstract

Many of the Android apps get published without appropriate security considerations, possibly due to not verifying code or not identifying vulnerabilities at the early stages of development. This can be overcome by using an AI based model trained on a properlly labeled dataset. Hence, LVDAndro provides a dataset for Android source code vulnerabilities, labelled based on Common Weakness Enumeration (CWE). The dataset has been generated using code lines scanned from real-world Android apps containing a large amount of distinct source code samples. The dataset can be downloaded from the Dataset directory. There are 3 dataset folders and each contains a readme file with important details and links to download dataset stored in a Google Drive.

Citation

SENANAYAKE, J., KALUTARAGE, H., AL-KADRI, M.O., PIRAS, L. and PETROVSKI, A. 2023. Labelled Vulnerability Dataset on Android source code (LVDAndro) to develop AI-based code vulnerability detection models [Dataset]. Hosted on GitHub (online). Available from: https://github.com/softwaresec-labs/LVDAndro

• Online Publication Date: Sep 2, 2022
• Publication Date: Sep 2, 2022
• Deposit Date: Sep 7, 2023
• Publicly Available Date: Sep 7, 2023
• Keywords: Android application security; Code vulnerability; Labelled dataset; Artificial intelligence; Auto machine learning
• Public URL: https://rgu-repository.worktribe.com/output/2072071
• Publisher URL: https://web.archive.org/web/20230907100039/https://github.com/softwaresec-labs/LVDAndro
• Related Public URLs: https://rgu-repository.worktribe.com/output/2072016 (Related conference paper)
• Collection Date: Sep 2, 2022