Broadening search parameters

[bcorrie]

Hello, I was just looking at the search parameters for your PMC query, and I was wondering if there was a reason for the particular scope that was used. It seems somewhat narrow on a couple of dimensions. I have a couple of suggestions to make this potentially more complete.

• It doesn't search for some common phrases that you find in the "Biodata resource" ecosystem, such as "repository" or "knowledge base"
• The URL matching relies on the URL having either the protocol listed (e.g. http) or assuming the URL has www in the Fully Qualified Domain Name (FQDN). I think in most cases www would indicate a web site and not a biodata resource. It is quite easy to create a regular expression for FQDNs that would capture any pattern with domain name pattern of x.y.z

Maybe a suggestion for the 2023 inventory 8-)

I do admit that I went to the 2022 Inventory and was wondering why I didn't find our resource and associated publication on the list - so I investigated why. 8-) Our paper is not listed I suspect because we used (gateway.ireceptor.org) which is clearly a domain name but has neither www or http.

Kudos in producing a data set and a process that could be reproduced so I could figure out why we didn't make the list!

[1heidi]

I am so sorry for not seeing this sooner, @bcorrie! Thank you for your interest and suggestions!

re: the query, yes, definitely, additional terms could be added. We were hoping that "data" would catch others besides databases and resources. It did but definitely not all, but we can use the alternate terms now as a sort of thesaurus, e.g. there were also quite a few "altas" an even some rarer ones like "index". I actually just finished up a little mini analysis of that. And this is also a great point about the URL. HTTP/WWW were obvious and are common, but you are correct that wouldn't catch all. not Another analysis could be all the various top level domains that were used so could catch those. Or moving outside of the query itself and working with the abstracts directly... so many directions. :)

I just saw iReceptor is registered in re3data - terrific! Hopefully someday we won't need to be doing these kinds of projects all! Thanks again for the feedback, and my apologies again for the delay!!