Open cessda-bitbucket-importer opened 2 years ago
Original comment by Kostas Papagiannopoulos (GitHub: kpapag).
we managed to score higher by resolving the “Namespaces of known semantic resources can be identified in metadata“. The link that needed to be updated is CESSDA Vocabulary Topic Classification", URI="https://vocabularies.cessda.eu/v2/vocabularies/TopicClassification/4.0?languageVersion=en-4.0 to the newest https://vocabularies.cessda.eu/vocabulary/TopicClassification?v=4.0
We have updated the vocabulary links first in the ElasticSearch record which did not change the score. However, when updating the links in OAI xml file in a test environment, the score had a small increase.
Example: https://datacatalogue.cessda.eu/detail?lang=en&q=ff5e71ee1ed3e426ea71ea048f8bc7af10c567de690085d94d50de6d59b1705b was initially scoring "FAIR": 43.75, now scores "FAIR": 47.92.
Need to contact Robert and see if FUJI can handle redirect mechanism, otherwise OAI files holding this link could be updated by the SP’s.
Original report on BitBucket by Kostas Papagiannopoulos (GitHub: kpapag).
After resolving the identifier issue, our FUJI results of studies that provide DOI are now about 45% FAIR compatible.
Posting here, for further review, the metrics that still fail and examine if something can be done from our side. Used this study as a test case, from localhost until identifier fix makes production:
https://datacatalogue.cessda.eu/detail?lang=en&q=fcdf8ad65e3650b4e3e8e5c379cdaaa67ee93b90addeb20616e9188c30e78db9