Coherence Across Components in Cognitive Systems — One Ontology to Rule Them All
Authors: Gregor Behnke, Denis Ponomaryov, Marvin Schiller, Pascal Bercher, Florian Nothdurft, Birte Glimm, Susanne Biundo
IJCAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In our case-study scenario the initial, non-extended planning domain contains 310 different tasks and only a few methods. The described ontology contains 1230 concepts (613 imported from the NCICB corpus) and 2903 axioms (of which 664 are from NCICB). This includes 310 concepts for integrating the planning domain into the ontology. Further, the ontology contains 9 different training objectives and 24 workout templates. The planning domain, expanded with new decompositions inferred from the ontology, contains 471 tasks and 967 methods. Our implemented system employs the OWL reasoner FaCT++ [Tsarkov and Horrocks, 2006]. On an up-to-date laptop computer (Intel R Core TM i5-4300U) it takes 3.6 seconds to compute the whole extended planning domain. Of the newly generated methods, 203 are created based upon workouts subsumed by workout templates and 3 methods have been created by combinations of concepts. Further, 59 decomposition methods for training objectives into workout templates have been found of which 24 are combinations of concepts. We would like to point out that every decomposition linking workouts and trainings in this scenario is inferred by the reasoner. |
| Researcher Affiliation | Academia | Institute of Artificial Intelligence, Ulm University, Germany A.P. Ershov Institute of Informatics Systems, Novosibirsk, Russia Institute of Communications Engineering, Ulm University, Germany |
| Pseudocode | No | The paper does not contain structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described, nor does it state that the code is publicly available. |
| Open Datasets | Yes | In our use-case the ontology already contains parts of the taxonomy of the NCICB corpus [NCICB, 2015], which describes muscles, joints and bones of the human body and their relations. ... [NCICB, 2015] NCI Center for Bioinformatics NCICB, 2015. http://ncicb.nci.nih.gov/xml/owl/EVS/Thesaurus. owl (accessed February 9, 2015). |
| Dataset Splits | No | The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning for machine learning models. |
| Hardware Specification | Yes | On an up-to-date laptop computer (Intel R Core TM i5-4300U) it takes 3.6 seconds to compute the whole extended planning domain. |
| Software Dependencies | No | Our implemented system employs the OWL reasoner FaCT++ [Tsarkov and Horrocks, 2006]. While a specific software (FaCT++) is mentioned and cited, its version number is not provided, nor are explicit version numbers for other relevant libraries or frameworks. |
| Experiment Setup | No | The paper describes the conceptual setup of their system and how knowledge is encoded in the ontology, but it does not provide specific experimental setup details such as concrete hyperparameter values, training configurations, or system-level settings typically found in machine learning experiments. |