Skip to content

Latest commit

 

History

History
67 lines (44 loc) · 2.77 KB

README.md

File metadata and controls

67 lines (44 loc) · 2.77 KB
Raw

Coverage Status NCSA

AgreementMaker Ontology Matching System

AgreementMaker is an ontology matching system was started by the ADVIS Laboratory at the University of Illinois at Chicago, under the supervision of Professor Isabel F. Cruz.

It has competed multiple times in the Ontology Alignment Evaluation Initiative and presented impressive results.

Project Organization

AgreementMaker is split into multiple projects, all in the projects sub-directory. To build Agreementmaker, use Gradle:

./gradlew build

Running

WordNet Dictionary Files

AgreementMaker requires the WordNet dictionary files at application startup. If it cannot find them it will show an error like this:

ERROR WordNetUtils java.lang.RuntimeException: Cannot find WordNet dictionary files. Expected the files to exist in this directory: /path/to/agreementmaker/AM_ROOT/wordnet-3.0

NOTE: On your computer the path /path/to/agreementmaker/AM_ROOT/wordnet-3.0 will be a real path.

You can ensure that the files are there by downloading them from the WordNet website (substitute /path/to/agreementmaker/AM_ROOT/wordnet-3.0 with the real path in the error message):

mkdir -p /path/to/agreementmaker/AM_ROOT/wordnet-3.0
cd /path/to/agreementmaker/AM_ROOT/wordnet-3.0
wget https://wordnetcode.princeton.edu/wn3.1.dict.tar.gz
tar --strip-components=1 -xvzf wn3.1.dict.tar.gz

To run AgreementMaker, use the run Gradle task:

./gradlew run

Sample Ontologies

We can use the OAEI 2012 ontology dataset for matching from the 2012 OAEI Campaign page.

  1. Download and unzip the benchmarks.zip.
  2. Start AgreementMaker with ./mvnw pax:provision.

In AgreementMaker:

  1. Navigate to File -> Open Ontologies... (Ctrl + O).
  2. In the Source Ontology tab, check the Load Ontology box, select the benchmarks/101/onto.rdf file, and set the language to OWL.
  3. In the Target Ontology tab, check the Load Ontology box, select the benchmarks/201/onto.rdf file, and set the language to OWL.
  4. Click Proceed, and Ok on the dialogs.

Now to match the ontologies:

  1. Click the Match! button in the Matching Tasks Control Panel.
  2. Notice that Base Similarity Matcher is selected in the Matching Algorithm tab.
  3. Click Run Matching Task to match the ontologies.