search

mastaal / nllegalcit

A Python library to find citations to Dutch legal documents in natural language text.
https://nllegalcit.martijn-staal.nl/latest/
European Union Public License 1.2
1 stars 0 forks source link

Compare performance with LiDO #13

Open mastaal opened 6 months ago

mastaal commented 6 months ago

LiDO has an API which can be used to systematically find the references their LinkeXtractor has found in a specific document, see https://linkeddata.overheid.nl/front/portal/services. This can be used to verify if nllegalcit finds at least the same citations in a given document.

mastaal commented 6 months ago

Groundwork for this - but only with regards to case law - is implemented in 6b553a621650058374ff10ade766f48e113d29a1

pascal1888 commented 2 months ago

Do note that the performance of LiDO will probably be significantly better because of the hardcoded strings of the LinkeXtractor, which are not implemented in this project right now (at point 679eeab).

See the first step in the LinkeXtractor pipeline (in Dutch): image

I will try to find time to contribute!