SemSys Papers Accepted in December
December wrapped up the year with a range of insightful research. From an approach for a new way judging the strength of arguments in formal discussions, pattern-based engineering of neurosymbolic AI systems to new tools for improving AI transparency, the following papers were accepted by SemSys authors:
One paper was accepted at the Thirty-Ninth AAAI Conference on Artificial Intelligence:
- L. Bengel, G. Buraglio, J. Maly and K. Skiba, An Extension-Based Argument-Ranking Semantics: Social Rankings in Abstract Argumentation
- Topic: The paper introduces a new approach to ranking arguments in abstract argumentation, a formal framework for representing discussions. This is achieved by using so-called social ranking functions which have been developed by economists to rank individuals based on their performance in groups.
Two papers were accepted at the Journal of Web Semantics Special Issue “Opportunities for Knowledge Graphs in the AI Landscape – An Application-Centric Perspective”:
- F. J. Ekaputra, Pattern-based engineering of Neurosymbolic AI Systems, Journal of Web Semantics, Volume 85, 2025.
- Topic: The paper outlines a vision for pattern-based approaches to engineer neuro-symbolic AI (NeSy-AI) systems. Using Knowledge Graphs at its core, the proposed approach connects visual and formal notations to streamline tasks like documentation and artefact generation, making the development process more efficient and traceable.
- L. Waltersdorfer, M. Sabou, Leveraging Knowledge Graphs for AI System Auditing and Transparency, Journal of Web Semantics, Volume 84, 2025.
- Topic: This work highlights limitations of current AI auditing processes and tools. The use of Knowledge Graphs is discussed for enabling higher auditability and transparency in AI systems.
One paper was accepted at AIGOV @AAAI 2025:
- L.Waltersdorfer D. Hausler and T. Auge: Provenance Question-based AI Transparency and Accountable AI Governance, 2025.
- Topic: This paper proposes an approach for using provenance questions as transparency requirements to translate them into executable queries for increased AI system transparency. Example questions are analysed on a linguistic and provenance level and a reference architecture is discussed.