I’ve recently been working with the Serials Solutions Summon API for some side work at George Washington University Libraries. Since the work is largely in Python I created a very simple reusable module called summoner for talking to the API. It largely just does the authentication bits, and gets out of the way…but it may be of interest to you if you use Python and work at a library that subscribes to Summon.

Anyway, I found myself wanting to get a picture of the metadata fields that are used in Summon responses, so I wrote a little script that does an open ended search and then walks through lots of results tallying up the fields used by content type. You can find the results for a scan of 50,000 records below.

As you can see the majority (78%) of the results were newspaper and journal articles. There was a fair bit of diversity in the fields returned for these. It was kind of interesting that 83% of the journal articles had ISSNs (I was expecting it to be higher). 50% had EISSNs so perhaps the 17% that lacked ISSNs had an EISSN instead. Only 41% of journals had subject terms. The sampling of the other content areas was probably too low to make any guesses.

The big caveat here is that these numbers reflect GWU’s particular holdings and subscription options. Still, if you use the Summon API you might find them interesting; and if you are really curious you can run the script for your own institution.

Newspaper Articles (54.0%)

Journal Articles (24.0%)

Reference (10.0%)

Patents (6.0%)

Books (2.0%)

Web Resources (2.0%)

Book Chapters (2.0%)