Ontology Matching

Graph Search Algorithm Based Prompt Generation for Ontology Matching.

Note: The repository consists of source codes of the paper "Exploring Prompt Generation Utilizing Graph Search Algorithms for Ontology Matching" presented at EU Semantics 2024. The paper can be accessed through the following link: ebooks.iospress.nl/doi/10.3233/SSW240003.

Folder Hierarchy

├── data
├── results
└── src
    ├── AlignmentFormat.py
    ├── accronyms.json
    ├── batch_loaders
    │   ├── alignment.py
    │   ├── ontology_parsing
    │   └── random_walk.py
    ├── config.json
    ├── configMatcher.json
    ├── configMatcherImport.py
    ├── globals.py
    ├── llms
    ├── maximum_bipartite_matching
    ├── prompt_template_generator
    ├── run_matcher.py
    ├── track.py
    ├── utilsODS.py
    └── verbalizer

Requirements

use Python version >=3.10

$ pip install -r requirements.txt

on macOS please run

$ brew install cmake
$ python3 -m nltk.downloader stopwords

Dataset Folder:

update dataset paths at src/config.json

Configuration:

adjust the pipeline tasks and algorithm configurations at src/configMatcher.json

How to run

$ git clone https://github.com/JulianSampels/OntoMatch.git

download this file and extract it to src/verbalizer/graph2text/outputs/t5-base_13881/

$ cd src
$ python3 run_matcher.py

Results

all results can be found at results/result_RDF
best results on the OAEI Conference track are at results/result_RDF/conference/treePromptVersion0

T-test

We conducted t-test with significance level 0.1 to evaluate the significance of the prompt types and algorithms. The results can be seen on t-test folder.