The Austrian Society for Artificial Intelligence (ASAI) has awarded Alexander Beiser its Master Thesis Prize 2026, one of two main prizes this year, for his diploma thesis Novel Techniques for Circumventing the ASP Bottleneck, supervised by Stefan Woltran and Markus Hecher. With the prize, which is supported by the Federal Ministry for Infrastructure, Mobility and Innovation (BMIMI), ASAI promotes young researchers by honoring the best diploma and master’s theses in a subfield of Artificial Intelligence.

The thesis tackles a central scalability problem of Answer Set Programming (ASP), the so-called grounding bottleneck. Replacing a program’s variables with all their possible values can blow it up exponentially, which leaves many industrial problems out of reach. Body-decoupled grounding (BDG) eases this bottleneck, but it could not be combined with traditional grounders and performed poorly on normal and non-tight programs. The thesis addresses both limitations. With hybrid grounding, BDG and traditional grounding each take over the parts of a program they handle best, and heuristics decide the split automatically. This solves about 35% more instances on grounding-heavy benchmarks while staying on par on solving-heavy ones. The new techniques FastFound and Lazy-BDG substantially improve BDG on normal and non-tight programs, outperforming both the original BDG and the state-of-the-art grounders gringo and idlv on synthetic benchmarks. The work formed the basis of three peer-reviewed publications.

The ASAI Prize is the latest in a series of distinctions for the thesis, following the Best Master Thesis Award of TU Wien Informatics and the Diplomarbeitspreis der Stadt Wien 2025. Alexander now continues his research as a PhD candidate at CAIML, developing AI methods for air traffic management in collaboration with Frequentis.

Links

• ASAI Prize: https://www.asai.ac.at/en/asai-prize
• Thesis (open access): https://doi.org/10.34726/hss.2025.127965