Page 1 of 1

PATSTAT Register: event codes & text

Posted: Mon Jan 29, 2018 4:11 pm
by EPO / PATSTAT Support
Quite some PATSTAT Register users have asked for extra information on the event codes and event text data which you can find in the reference table REG402_EVENT_TEXT .

Factum is that we do not have much extra information then wat is available in the event_text field. But it might look a bit strange when you carry out a query (below) to simply list the unique (distinct) event_code and event_text attributes; you will find different codes with the same text. Why ?

Code: Select all

SELECT distinct event_code, event_text
  FROM reg402_event_text
  order by event_text
These codes are being used by our back-end systems to identify mainly the "time" the event took place from a point of view of the granting procedure. This embedded “extra” information that does not really effect the meaning of the event itself.
For example:

Times the even took place embedded in the code:
- 9299 in the code means : comes up after B1 publication
- 9199 in the code means : comes up before B1 publication
- 8199 in the code means : you can ignore it, not implemented (yet)

- Having “EPIDOS” in the codes means that the event will be triggering a daily update of the data (so processing time is more or less real in for example EPO Register); other event codes are weekly updates (batch processes.

Other embedded codes:
- C – Change (e.g.: EPIDOSCRFE2)
- N – New (e.g.: EPIDOSNRFE2)
- D – Delete (e.g.: EPIDOSDRFE2)
As the amount of codes is rathter big (545), it is rather difficult for the analyst and researcher to get a proper view on the "importance" of a code. To make this easier, we have made an excel sheet that counts for each year of application the number of times a certain event_code was used. This illustrates immeditely that large amount of the codes have been used very few times. (and create a lot of noice in analysis). It also shows that a large of amount of codes were introduced and discontinued in 2002-2003. We do not have further information for the time being, but contributions from the community are welcomed.

Geert BOEDT