Data used to evaluate ORBITS: Optimal Repair-Based Inconsistency-Tolerant Semantics

This dataset provides the input files that were used in the evaluation of the ORBITS system (Optimal Repair-Based Inconsistency-Tolerant Semantics, https://github.com/bourgaux/orbits). A detailed description is available in a technical report on arXiv (https://arxiv.org/abs/2202.07980).

Content:

Folders cqapri_benchmark, food_inspection_benchmark, and physicians_benchmark contain JSON files of conflict graphs and candidate queries and their causes.
These files are named using the following pattern: files of candidate answers and their causes are named __answers_causes.json, and conflict graphs are named _conflictGraph_.json where says whether the priority relation is score-structured (prio_score) or not (prio_non_score) and the probability (p) or number of scores (n) used to build the priority relation.

Folder original_datasets_and_queries contains the Food Inspection and Physicians datasets used to generate files from food_inspection_benchmark and physicians_benchmark.
Files from cqapri_benchmark have been generated from the CQAPri benchmark available at https://lahdak.lri.fr/CQAPri/CQAPri.php.
In all cases, we use ProvSQL (https://github.com/PierreSenellart/provsql) to build conflict graphs and causes from the datasets.