Query-Driven Repairing of Inconsistent DL-Lite Knowledge Bases

Authors: Meghyn Bienvenu, Camille Bourgaux, François Goasdoué

IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We report on experiments made on core components of the above Opt DRP algorithm. We focused on measuring the time to decide whether a potential solution exists (Step B), to compute necessarily (non)false and relevant assertions (Step B.1), to rank the relevant assertions w.r.t. their impact (Step B.2.c), and to find the MCSWs (Step C).
Researcher Affiliation Academia Meghyn Bienvenu CNRS, Univ. Montpellier, Inria Montpellier, France Camille Bourgaux Univ. Paris-Sud, CNRS Orsay, France Franc ois Goasdou e Univ. Rennes 1, CNRS Lannion, France
Pseudocode Yes Figure 1: Algorithm for constructing a globally U and locally {U,W}-optimal repair plan
Open Source Code No The paper provides links to third-party tools (SAT4J, CQAPri) but does not state that the code for their own methodology is open-source or provide a link to their own implementation's repository.
Open Datasets Yes We borrowed from the CQAPri benchmark [Bienvenu et al., 2016a] available at the URL above its: (i) TBox which is the DL-Lite R version of the Lehigh University Benchmark [Lutz et al., 2013] augmented with constraints allowing for conflicts, (ii) c5 and c29 ABoxes with 10 million assertions and, respectively, a ratio of assertions involved in conflicts of 5%, that we found realistic, and of 29%.
Dataset Splits No The paper describes the datasets used (c5 and c29 ABoxes) but does not specify how they were split into training, validation, or test sets.
Hardware Specification No The paper does not provide any specific hardware details such as CPU or GPU models, or cloud computing specifications used for running the experiments.
Software Dependencies No The paper mentions developing components in Java and using the CQAPri system and SAT4J, but it does not provide specific version numbers for these software dependencies (e.g., 'Java 7', 'SAT4J 2.0').
Experiment Setup No The paper describes the algorithms and measures their computational time but does not detail specific experimental setup parameters such as hyperparameters, training configurations, or model-specific settings.