By John Holmberg Runsten
This tutorial explains how OpenRefine (http://openrefine.org/) together with the oaDOI API (https://oadoi.org/) from Impactstory can be used to collect Open Access information. aoDOI collects open access evidence from DOAJ, crossref, BASE, and pmicd. This can be useful when trying to identify OA articles published by one’s institution. In this tutorial records from Scopus have been used.
Collecting-Open-Access-information-using-OpenRefine-and-the-oaDOI-API
Currently the BASE data has a problem which has an effect on identifying if an article is available from institutional repositories and this leads to false negatives.
It seems that BASE is using the
content of first dc:rights element in the data to determine the OA status
of the harvested item. It is quite common that dc:rights is being repeated
in the metadata with varying values (as it is mentioned in the documentation
of BASE OAI) and when the info:eu-repo term for indicating OA status is
not in the first available dc:rights field, the OA status of the item is
marked Unknown in the BASE database and therefore are not being found via OADOI.
Good to know Matti!
Concerning the accuracy of oaDOI this study (https://peerj.com/preprints/3119v1/) states that “77% of the truly open articles are correctly identified as open by oaDO […and] 96.6% of the time that oaDOI reports an
article is open, it really is open.”
The study is however not yet reviewed.
It seems that this problem in the Base data has been resolved recently as the articles I have been using as a test case resolve correctly to our repository via OADOI.