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 and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Querying Inconsistent Description Logic Knowledge Bases under Preferred Repair Semantics
Authors: Meghyn Bienvenu, Camille Bourgaux, François Goasdoué
AAAI 2014 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | An experimental evaluation of the approach shows good scalability on realistic cases. Our second contribution is a practical approach to query answering in DL-Lite R under the AR, P -AR, and P -IAR semantics... An experimental evaluation demonstrates the scalability of the approach in settings we presume realistic. 6 Experimental Evaluation We implemented our query answering framework in Java within our CQAPri ( Consistent Query Answering with Priorities ) tool. |
| Researcher Affiliation | Academia | Meghyn Bienvenu and Camille Bourgaux LRI, CNRS & Universit e Paris-Sud Orsay, France Franc ois Goasdou e IRISA, Universit e de Rennes 1 Lannion, France |
| Pseudocode | No | No pseudocode or algorithm blocks are explicitly presented. |
| Open Source Code | No | We implemented our query answering framework in Java within our CQAPri ( Consistent Query Answering with Priorities ) tool. CQAPri is built on top of the relational database server Postgre SQL, the Rapid query rewriting engine for DL-Lite (Chortaras, Trivela, and Stamou 2011), and the SAT4J v2.3.4 SAT solver (Berre and Parrain 2010). |
| Open Datasets | Yes | We considered the modified LUBM benchmark from Lutz et al. (2013), which provides the DL-Lite R version LUBM 20 of the original LUBM ELI TBox, and the Extended University Data Generator (EUDG) v0.1a (both available at www.informatik.uni-bremen.de/ clu/combined). |
| Dataset Splits | No | Inconsistencies in the ABox were introduced by contradicting the presence of an individual in a concept assertion with probability p, and the presence of each individual in a role assertion with probability p/2. Additionally, for every role assertion, its individuals are switched with probability p/10. Prioritizations of ABox were made to capture a variety of scenarios. |
| Hardware Specification | No | No specific hardware details (like CPU/GPU models, memory, etc.) are mentioned for running experiments. |
| Software Dependencies | Yes | CQAPri is built on top of the relational database server Postgre SQL, the Rapid query rewriting engine for DL-Lite (Chortaras, Trivela, and Stamou 2011), and the SAT4J v2.3.4 SAT solver (Berre and Parrain 2010). |
| Experiment Setup | Yes | Inconsistencies in the ABox were introduced by contradicting the presence of an individual in a concept assertion with probability p, and the presence of each individual in a role assertion with probability p/2. Additionally, for every role assertion, its individuals are switched with probability p/10. Prioritizations of ABox were made to capture a variety of scenarios... Every ABox s id u Xp Y indicates the number X of universities generated by EUDG and the probability value Y of p for adding inconsistencies as explained above (Me-P reads M.10 P ).... We built 8 prioritizations for each of these ABoxes further denoted by the id of the ABox it derives from, and a suffix first indicating the number of priority levels and then how these levels were chosen. l Zd W indicates the number Z of priority levels: 3 and 10 in our experiments, and the distribution W: cr=, a=, cr =, or a = indicates whether priority levels were chosen per concept/role (cr) or assertion (a), and whether choosing between these levels was equiprobable (=) or not (=). |