Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]
Embedding Ontologies in the Description Logic ALC by Axis-Aligned Cones
Authors: Özgür Lütfü Özcep, Mena Leemhuis, Diedrich Wolter
JAIR 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | Our main result states that an ALC ontology is satisfiable in a classical sense iff it is satisfiable by a geometric model that interprets all concept descriptions as cones (concretely: axis-aligned cones). We derive this result by first considering the Boolean part of ALC ontologies and then generalizing it to full ALC. The geometric models we use are partial and thus allow some uncertainty to be retained, i.e., if x is only known to be a member of the union of two atomic concepts, then our partial model will not commit to saying to which atomic concept x belongs. ... Development of a practical learning algorithm and its experimental evaluation are outside the scope of this paper. |
| Researcher Affiliation | Academia | Ozg ur L utf u Oz cep EMAIL Mena Leemhuis EMAIL University of L ubeck, Germany Diedrich Wolter EMAIL University of Bamberg, Germany |
| Pseudocode | No | The paper primarily focuses on theoretical developments, including definitions, propositions, lemmas, and proofs related to embedding ontologies in description logic. It does not contain any clearly labeled pseudocode or algorithm blocks, nor does it present structured steps in a code-like format. |
| Open Source Code | No | Development of a practical learning algorithm and its experimental evaluation are outside the scope of this paper. Shortly after Oz cep et al. (2020) presented the idea of cone-based geometric models for the first time, learning methods using cones appeared in Neur IPS papers (Zhang, Wang, Jiajun, Shuiwang, & Feng, 2021; Bai, Ying, Ren, & Leskovec, 2021). However, these are not aligned with logic methods. |
| Open Datasets | No | The paper does not conduct experiments and therefore does not use any datasets. It mentions datasets such as CIFAR-10 in the context of related work by other researchers, but not as part of its own methodology or evaluation. |
| Dataset Splits | No | The paper is theoretical and does not involve experimental evaluation using datasets. Therefore, there is no mention of training/test/validation dataset splits. |
| Hardware Specification | No | The paper is theoretical and does not report on experimental results requiring specific hardware. The text explicitly states: 'Development of a practical learning algorithm and its experimental evaluation are outside the scope of this paper.' |
| Software Dependencies | No | The paper is theoretical and does not describe any practical implementation or experimental setup that would require specific software dependencies with version numbers. |
| Experiment Setup | No | The paper is theoretical, focusing on mathematical and logical foundations. It explicitly states that 'Development of a practical learning algorithm and its experimental evaluation are outside the scope of this paper,' and thus does not provide experimental setup details, hyperparameters, or training configurations. |