EML search queries about eml HOT 4 CLOSED

@cboettig -- yeah, for searching across documents, I think we really should be using another library like the R dataone library that let's us do full SOLR queries against EML and other metadata standards using any repository that is DataONE enabled. Querying across documents requires indexing those documents as a collection, and that seems out of scope for reml. The dataone library might be used like this:

library(dataone)

# Initialize a client to interact with DataONE
cli <- D1Client()

# search for data packages and only request versions that are not obsolete
results <- d1SolrQuery(cli, list(q="increase experiments -obsoletedBy:*",fl="identifier,title,author,documents,resourceMap"))
sr <- xmlParse(results)
sr

There are hundreds of thousands of EML documents and FGDC documents accessible via the DataONE search index.

from eml.

@mbjones This sounds great.

I think one useful exercise in this spirit would simply be to give a snapshot of what is available in EML across, say, all of DataONE: how many EML files, covering how many publications, how many taxa (or distribution of coverage across taxa), how many attributes, rows of data; what fraction are from LTER sites, fraction that are EML vs FGDC, number of EML files by year, etc. I think many researchers might be surprised to realize how much is available, which could help motivate further adoption.

from eml.

This would be particularly important to emphasize for the manuscript to really drive home the point. Ecologist (as with most scientists) are reluctant to invest time in a tool like this without knowing what their time investment is really worth.

from eml.

Closing this issue since SOLR queries are part of dataone and external to reml.

We can nonetheless include a sample solr query in the example documentation showing uploading to the KNB, illustrating that the EML metadata submitted becomes a query-able part of the dataone general index. After all, that fact alone is a huge advantage of using EML (+KNB) and important to bring home to folks. (vs using any other metadata standard or depositing on Dryad or Figshare where only much more minimal metadata is exposed).

from eml.