Hello José,
your query is correct; but a couple of small remarks.
Generally spoken, if you really want to limit to applicants, you should force the applt_seq_nr to be > 0 . In your query it does not make any difference because of the conditions on the person_name. Also, as you are looking for publications applied by Spanish Universities, one could make use of PSN_SECTOR attribute to retrieve possible "university" applicants that do not fulfil the like 'universi%' condition. The code below could be added in your query.
Code: Select all
AND (person_name like 'universi%' or psn_sector in ('UNIVERSITY','UNIVERSITY HOSPITAL'))
AND applt_seq_nr > 0
Your query will result in 101 applications. It is important to remember that PATSTAT is produced only 2 times/year, and will therefore never show the "very latest" of the available data. If you then limit to publications in 2017, this restriction really limits the result and is not representative. The SPRING 2017 PATSTAT version is produced on the backfile extraction January 2017, any you can see from the result that all those publications are dated in January. So this is not so good.
Specific for your case; to get more up-to-date results I would use the GPI (Global Patent Index), which is weekly updated.
Approach: I extracted in PATSTAT all the "DOC_STD_NAME" occurrences for Spanish universities. Based on the PATSTAT list (which included a couple of false positives which I cleaned out), I then used excel to create the GPI query (which is not SQL based), followed by a search and extraction in GPI. This results in 962 documents. Data on those documents were then extracted as an excel sheet. (In the GPI preferences, there are several options on what data fields can be extracted.) Further data aggregation in GPI is more complicated, but if the purpose is simply a list, then I think this will suit your needs. You can equaly download the PDF's.
Attached is the excel sheet with the names from PATSTAT, GPI query and results.