All notable changes to this project will be documented in this file. This project adheres to Semantic Versioning.
- Restructured the package so it can be easily imported and used as a library.
- Added setup.py
- Added support for simple (CSV) inputs.
- Filter out anonymous RDF graph nodes.
- Updated
rdflib
version.
- Added
none
statistical adjustment option.
- Added option
-O | --optimalsubclass
for finding the optimal subclass at each specialization step.
- Added support for custom annotation naming in rules. This allows the user to give meaningful names to annotation relationships. For example, in
example/finance/data.n3
we define a custom annotation name as follows:
ns1:GeographicalRegion
ns2:annotation_name "geo_region" .
Concepts that are descendants of ns1:GeographicalRegion
will appear in rules using the geo_region
name instead of the default annotated_with
predicate.
- Major changes in how candidate rules are evaluated.