Description
Announcement: Thesis (MA/BA) project in cooperation with Fraunhofer AISEC
Solve the points-to problem with ML
This project can be realized as a master or bachelor thesis
Motivation and Task Description
In static code analysis, the points-to problem involves determining the allocation sites to which poin-
ters may refer. Due to its undecidability in static contexts, only approximate solutions are feasible.
Existing methods often over-approximate, leading to numerous false positives. In this project we want
to leverage machine learning to achieve a more accurate approximation of the points-to problem.
From previous work an existing dataset derived from dynamic analysis can be used and extended as
needed. The core challenge lies in designing a ML model and preparing the input data appropriately.
As a result, the model should predict the likelihood that a pointer refers to a specific allocation site.
Requirements
• Programming skills: Python, ideally C/C++
• Practical experience with ML and PyTourch (ideally Graph Neuronal Networks)
• Knowledge about AST and control flow graph
• High amount of self motivation and independent work
Contact
Hannah Schmid, Tobias Specht
Telefon: +49 89 322-9986-130, Telefon: +49 89 322-9986-187
E-Mail: hannah.schmid@aisec.fraunhofer.de
E-Mail: tobias.specht@aisec.fraunhofer.de
Fraunhofer Research Institution for Applied and Integrated Security (AISEC)
Product Protection and Industrial Security
Lichtenbergstraße 11, 85748 Garching (near Munich), Germany
https://www.aisec.fraunhofer.de
|
Thesis topic in co-operation with the Fraunhofer Institute for Applied and Integrated Security AISEC, Garching