Acknowledgements

This tool is inspired from previous works in the literature:

• On the Measure of Conflicts: Shapley Inconsistency Values. Hunter and Konieczny (2010).
• Inconsistency Measures for Repair Semantics in OBDA. Yun et al (2018).
• DLGP: An extended Datalog Syntax for Existential Rules and Datalog±. GraphIK Team∗, LIRMM (2017)
• Graal: A Toolkit for Query Answering with Existential Rules. Baget, M. Leclère, M.-L. Mugnier, S. Rocher, and C. Sipieter (2015)

Description

This tool is designed for decision-making using three types of inconsistency measures (and values) from an inconsistent knowledge base.

The input of the tool are:

• A set of options in debate
• A context that describes the characteristics of the options
• A knowledge base for each user composed of preference statements and rules.

The output of the tool is:

• The score of each element of the user knowledge base w.r.t. the different Shapley inconsistency values.
• (Upcoming feature) A consistent partial preference for each user that is computing using a Shapley inconsistency value.

Usage

The tool should be launched using the following command: java -jar preference_resolution.jar [Option file] [Context file] [Preference file] where:

• [Option file] is the file containing the options. The options should be given one per line and should follow the constant notation of the DLGP format.
• [Context file] is the file containing the atoms describing the characteristics of the options. They should be ground atoms using only options and follow the fact notation of the DLGP format.
• [Preference file] is the file containing the preference rules/atoms given by the users.

Example:

java -jar preference_resolution.jar Examples/AlgorithmSelection/options.txt Examples/AlgorithmSelection/context.dlgp Examples/AlgorithmSelection/preferences.dlgp

Content

This repository contains:

• The source-code of the tool
• A compiled jar of the tool
• Simple examples

Contacts

In order to contact me, send me an email at: yun@lirmm.fr