EPO - Discussion forums

I would like to replicate the results reported in the Patent Index 2020 - European patent applications file (https://www.epo.org/about-us/annual-rep ... stics.html) for the indicator 'Total applications - split by main countries' in relation to the total of patents applied between 2015 and 2021. My Query shows results that are representatively higher than the values disclosed by the EPO, possibly due to duplicate values or inadequate code.

Did a colleague try to do this data collection?

Thanks,
Fernando

Dear Fernando,
which EPO product(s) are you using? GPI? PATSTAT? ..
It would help us if you could share your query with us.
Thank you and best regards,
epal team

Dear EPO team,
I am using specifically PATSTAT EP register database.
I propose to get a list of all the data in the patent application. Are expected that the sum of individual lines matches the sum value by country corresponding to the total value posted on the spreadsheet by the EPO. So I will do the same for the sectors.
Thank you and best regards,
Fernando

Dear Fernando,
using SQL in PATSTAT Register, there is indeed a risk for double counting when multiple tables are joined in a query. Each joined table might multiplicate the number of rows if no further restrictions are applied in the WHERE clause.

Generally spoken on using PATSTAT for "replicating" figures published by the EPO,have a look at this PATSTAT forum topic:
count-number-of-patent-applications-8372#p24162

So normally, the application figures published in the official EPO statistics will be about 10% higher then what you can find via PATSTAT Global or PATSTAT Register. Without seeing your query, it is not possible to provide more thoughts on the possible reason for the differences.