An Application For Anonymizing Ontologies.
OntoAnon iterates through all triples of a graph and anonymizes/removes the data while preserving the ontology structure. Data is preserved by keeping standard vocabulary like the ontology languages OWL, RDF(S), SHACL, but also standardized ontologies like FOAF, DC, SKOS. A full list of kept namespaces can be found here. OntoAnon runs locally, so no data leaves the computer. It allows sharing the structural data and using online tools like neontometrics or the OntolOlogy Pitfall Scanner (OOPS), without giving away company secrets.
OntoAnon builds on the python GUI TKinter and rdflib. The former is preinstalled with current python versions, the latter can be acquired using pip or the corresponding pipfile.
The tool is part of a publication currently under review. The development of the tool was part of a bachelor thesis by Robert Schlücker, cf. https://github.com/Junech1/Ontology-Anonymisation for the original repository.
%%{init: {'theme':'neutral', fontSize: '20px'}}%%
graph LR
A([Start]) --> D
subgraph Iterate over elements
C -- True:<br>Go to <br>next element --> D{Check if<br>standard<br> namespace}
D -- True --> E[Do not translate]
I -- True --> J{Check if<br>object is<br>already<br> translated}
J -- True --> K[Use<br>existing<br>translation]
K --> F
S -- Is blind-node -->
J -- False --> M[Translate with <br>counter variable]
M --> F
I -- False --> O[Translate<br>namespace]
O --> M
D -- False --> S{Check<br>element<br>type}
S -- Is<br>URI --> I{Check if<br>URI is in<br> a known<br>namespace}
S -- Is<bR>literal --> T{Check if element<br> is a number}
T -- True --> E
T --> M
E --> F[counter + 1]
F --> C{More<br>Elements <br> available?}
end
C -- False --> AB([Return<br>translation])
```