| Identifier |
Short Explanation |
| A |
Each concept not only gets automatically its labels and descriptions in four languages, but for a mathematically correct definition all its superclass/type information from DBpedia,too. On one hand, that brings a broader view and more interconnections. On the other hand, it makes the graph much more complex. Today, I'm not sure whether it was really a good idea to add/generate the type information for each concept. |
| B |
Use only one base language (e.g. English) for the generated descriptions; this reduces the size of the models. Use ad-hoc Natural Language Processing (NLP) / Machine Learning (ML) to do ad-hoc translations. |
| C |
Because of maintenance and structural reasons, it would have been better to organize each ontology layer in an own RDFs/OWL-GRAPH element and keep it not only in a triple store, but in a quad store. |
| D |
Because of financial and experimental reasons (Cloud Computing), I selectively added the concept definitions (term = superclasses AND specific concept properties) by generating RDFs/OWL-statements from DBpedia. For a productive environment, it would be much nicer and more dynamic to use the Linked Open Data Cloud online. |
| E |
In which respect would JSON-LD simplify the model representation?
|
| F |
...
|