I was recently asked to find equivalents to eight Canadian patent
applications, and found patent families containing EP, WO etc for five of
these by using a combination of the WPI, Inpadoc and HCA files on STN.
However patent family information for the remaining three applications
proved elusive, and I did not fine them by searching the Dialog versions of
these files, Pluspat on Questel, and the patent number or application fields
on Espacenet. In each case the CA patent applications either side of the
required numbers was present when expanding the relevant indices for the
numbers.
I found the three CA numbers on the Canadian Patent office web site, and
these quoted the PCT application and publication numbers, one case also
quoted the UK application numbers a second case gave the US application
number.
I was by now curious as to how the Canadian patent office found the
equivalents, whereas Inpadoc had apparently not made the connection, and
Derwent and Chemical Abstracts had not picked them up either. I submitted
one number to STN to investigate and this produced some interesting results.
1. The case I quoted turns out to be a PCT transfer, which is not covered by
WPI and will only appear if and when granted, as an equivalent to the PCT
application.
2. The Inpadoc file on STN does list the CA application, but not in the /PN
field, which I searched, but rather in the /LSPN field. Hence you need to be
aware of either the PCT number obviously searchable in the normal /PN field
or more appropriately should always search CA numbers in both /PN and /LSPN
fields on STN.
However a response to STN from Inpadoc confirmed that they rely on National
Patent Offices for data, and this suggested that the Canadian data is
incomplete which means there is no way of checking the serial numbers.
3. Chemical Abstracts located the CA patent application number, identified
it as an equivalent and added it to their file as a result of my enquiry via
STN.
I was intrigued by these findings and therefore conducted a subject search
limited to the period 1997-2002 on the Canadian PTO Website, selected the
patent numbers and searched them in the three STN files to determine the hit
rate for CA patent applications in these files.
The results were as follows:-
Total number of CA numbers selected from CIPO search 47 ---- WPI 4 hits,
HCA 16 hits, Inpadoc 14 hits (38 hits using a combination of /PN and
/LSPN).
This whole study raises some questions in my mind,
Is it essential / desirable to identify CA applications and their
equivalents at an early stage?
* If so, what is the most effective method?
* In the above example -- why are the remaining 9 hits from CIPO not
in Inpadoc, and does this reflect incomplete provision of data by CIPO to
Inpadoc or some other factors?
* Should Derwent and Chemical Abstracts be encouraged to pursue early
stage CA applications more systematically?
* Can CA patent applications be searched for in Espacenet given the
above findings for Inpadoc?
* How does the Canadian PTO find equivalents to its applications?
* How are CA numbers released to the "patent public", such that Patent
Attorneys pick up these numbers for an invention, but it appears that they
do not always first see the EP, PCT, GB or US equivalents?
My apologies for the length and detail of this message, but the
subject intrigued me, and I look forward to the responses of my expert
colleagues around the globe.
Best Regards
Neil Walden
Patent Information Consultant
Information Management PGRDi Sandwich
Tel: +44 1304 642351 Fax: +44 1304 658300
E-mail : Neil_Walden@sandwich.pfizer.com
----------------------------------------------------------------
This message and any attachment has been virus checked by
Pfizer Corporate Information Technology, Sandwich.
----------------------------------------------------------------
This archive was generated by hypermail 2b30 : Fri Feb 14 2003 - 11:56:50 GMT