representative document of patent family

Here you can post your opinions, ask questions and share experiences on the PATSTAT product line. Please always indicate the PATSTAT edition (e.g. 2015 Autumn Edition) and the database (e.g. PATSTAT Online, MySQL, MS SQL Server, ...) you are using.
Post Reply

bjurgens
Posts: 12
Joined: Fri Jul 27, 2012 9:24 am

representative document of patent family

Post by bjurgens » Thu Dec 22, 2016 1:55 pm

Dear PATSTAT Team

I was wondering if PATSTAT has a way to identify a representative document of the patent family to which the selected document belongs when displaying result lists, using the rule set you use for ESPACENET (=> "Espacenet reference documents" in: https://www.epo.org/searching-for-paten ... cenet.html )

If not, do you have any suggestions how to implement it using PATSTAT?

Thanks & Happy Xmas!

Björn Jürgens


mkracker
Posts: 120
Joined: Wed Sep 04, 2013 6:17 am
Location: Vienna

Re: representative document of patent family

Post by mkracker » Tue Jan 10, 2017 8:51 am

Hi Björn,

No, there is no definition of a "representative publication" in PATSTAT. But you can define your own representative, using your own rules. These rules may be similar to the ones in Espacenet or totally different, whatever best suits your needs.

I would do it like this:
  1. Add 2 columns to the table which contains publications, like your result table or TLS211_PAT_PUBLN.
    Insert a score value in one of the new columns. Add a higher score for publications which are more "representative". You may use rules like "If it is the first EP publication with an English abstract, then score will be 10. If it is the first publication of an US or GB application, then score will be 8. If ...".
    Be sure that there is a default value, so you cover all records.
  2. The second column should have one of two values, like Y if it is a representative publication and N if it is not.
    Compute this value by assigning a Y to the publication of a family with the highest score, and N else.
    Alternatively, if you are only interested in the representative publications, you remove all publications which do not have the highest representative score of a family.
    If multiple publications of a family are equally representative, take any one or define another rule to distinguish between them.
This can be easily be done if you have PATSTAT Raw Data and you have loaded it into your own database.
In PATSTAT Online you cannot add or change columns, but you can download your result table and compute the representative publication offline with Excel or whatever tool you prefer in a similar way.

Have success!
Martin, EPO / PATSTAT
-------------------------------------------
Martin Kracker / EPO


Post Reply