1 (edited by Darwyn Sumner 08-02-2018 10:16:37)

Topic: Article on trust in public databases

This is worth a read:
Maldonado C, Molina CI, Zizka A, Persson C, Taylor CM, et al. 2015. Estimating species diversity and distribution in the era of Big Data: To what extent can we trust public databases? Glob. Ecol. Biogeogr. 24(8):973–84
(download from http://onlinelibrary.wiley.com/doi/10.1 … 6/abstract
Though the analyses are South American, some of the observations are highly relevant, I picked up on

Ideally, careful quality evaluation of the primary information in a dataset downloaded from GBIF should be conducted, before the data is used for further analyses – in particular checking, where possible, the species identification and locality data, including their georeferences (coordinates for latitude and longitude)

and the comments about data validation and new GBIF initiatives to improve it (our Atlas data feeds into GBIF so the same comments apply).
For workers like me trying to pull georeferences out of published papers across Europe or museum workers trying to do the same from specimen labels their comments about georeferencing errors are valuable.
There's a reference too about a topic raised recently on this forum (https://forums.nbn.org.uk/viewtopic.php?id=7134)

It is unfortunate that GBIF still do not allow their users to easily provide feedback on specific records, e.g. correcting misidentifications and erroneous georeferencing

All in all it is very positive about Big Data and is a highly quotable paper.

2

Re: Article on trust in public databases

Jo Judge told me that there might be some chance that NBN would develop mechanisms for workers who build up species occurrence datasets on mainland Europe to have their records submitted to GBIF. NBN is the GBIF "node" tasked to do that. Mike Weideli tells me there are some other R6 users who use it for European records.
Any news on progress with that?