Monday, December 10, 2012

A few useful links about trademarks

A trade mark is a word, phrase or symbol, or a combination of words, phrases, symbols or figurative elements, that identifies and distinguishes the goods or services of one company from those of another. Organisations ranging from SMEs to huge multinationals use trade marks registration systems to protect their brands.
 
These is a first list of interesting links about trademarks

USPTO Bulk Downloads: Trademarks - Google repository for US TM raw data 
http://www.google.com/googlebooks/uspto-trademarks.html

OECD Working paper about trademarks as an indicator of product and marketing innovations
http://www.oecd.org/sti/innovationinsciencetechnologyandindustry/42534274.pdf

EU TM office search page:

http://www.tmview.europa.eu/tmview/welcome.html

MARKIFY: alternative TM search engine for EU

http://trademark.markify.com/trademarks/ctm/ariminolva/001999580
 

Tuesday, December 4, 2012

IPC - SITC/ISIC concordance

 Lybber and Zolas present in the paper:

http://www.oecd.org/site/stipatents/1-2-Lybbert.pdf

A new algorithmic approach to constructing concordances between the International Patent Classification (IPC) system that organizes patents by technical features and industry classification systems that organize economic data, such as the Standard International Trade Classification (SITC), the International Standard Industrial Classification (ISIC) and the Harmonized System (HS). This ‘Algorithmic Links with Probabilities’ (ALP) approach incorporates text analysis software and keyword extraction programs and applies them to a comprehensive patent dataset.

The ALP concordances described in this paper can be downloaded from the WIPO website at http://www.wipo.int/econ_stat/en/economics/publications.html

About patent value



This is a short list about patent value speaker, at session held during PATSTAT conference 2012

Takalo:
Valuation of innovation: The case of iPhone


Other non academic participant were:

ocean tomo:
A US company who patented an algorithm to predict likely value of a patent;

designed for objectively assessing patent quality, relative value, relevant patents and technologies, competition, and competitive trends.

Creator af an algorithm called FICO for evaluating patents using citation, in a way similar to googlerank.
more at: http://www.crie.org/web/crie/patent-data/patent-rank/
http://www.youtube.com/watch?v=hxlBON6ebr0

 

UK IP office data interface


A good source of already structured data is also UK ip office.

In it's IPSUM (Online Patent Information and Document Inspection Service) interface allows users to access (and easily parse) many data, including EP reister data.


Data available are:

Application Number, Application Source, Application Language, Publication Number, Publication Language, Status (ie Granted), Filing Date, Publication Date, Grant Date, Last Renewal Date, Year of Last Renewal, Next Renewal Date, Designated States, Application Title. Grant Title, Address for Service + adp, Applicant / Proprietor + adp, Inventors + adp, EPO Representative + adp


IE:http://www.ipo.gov.uk/p-ipsum/Case/PublicationNumber/EP1000000

ADP stands for automated data processing number that is like person_id in Patstat, that unfortunately is not disambiguated (see new holland example below).



Applicant / Proprietor NEW HOLLAND KOBELCO CONSTRUCTION MACHINERY S.P.A. Strada di Settimo, 323 10099 San Mauro Torinese Italy [ADP Number 77324077001]

plicant / Proprietor NEW HOLLAND KOBELCO CONSTRUCTION MACHINERY S.P.A. Strada di Settimo, 323 10099 San Mauro Torinese (TO) Italy [ADP Number 76169127001]

Friday, November 30, 2012

News from patstat conference 2012


At 2012 OECD guested ediction of patstat conference people from EPO staff announced new developments for patstat db for coming months.

Most relevant are:

TLS217, cointaining eclas, will become tls224 since cpc will replace ecla.
Cpc (achronim of cooperative patent classification, developed jointly from EPO and USPTO) is an evolution of ecla extending it upto 250.000 entries. more on:
www.cpcinfo.org

Another change will affect inpadoc families table TLS219 where replenished (artificial) publications will be removed.
In TLS211 replenished publications ids will be in 900.0000.000 series.
Fields appln_id, publn_id will be changed by removing leading spaces.

Citations table quality will be improved by adding jp kr cn citations in 2013 (nowadays data exist but with a lot of gaps).

Espacenet will be enriched with south american and arabic patent authorities (also Morocco, Jordan...) and soon will be transferred into patstat.

As new product, eu register data (
http://www.epo.org/searching/free/register.html) will be distributed in Csv starting from 1st quarter 2013 as beta public test.
Since october 2013 data public distribution will start. Data size should be 22 gb, 2,5 gb zipped.





Some news arrived also about unitarypatent: the goal is to have the first grant in 2014, it will cohexist with EP patents and very likely at data level will have same application and pubblication authority (ep) but probably a new appln kind.
 
Eventually, t has been rumored that in a fortcoming ediction orbis (
http://www.bvdinfo.com/products/company-information/international/orbis) will give link to patstat.

Friday, November 23, 2012

Patent claims in Patstat

Recently available, the count of patent claims in TLS211 in patstat, looks like a very promising piece of information.

Coverage of such data is quite good, here the counts for the two publication authorities covered:
 
EP
2669148
US
4574142

About time coverage here are the figures for last 30 yrs in percentage (consider obviously that some publication kinds do not have a figure for claims) on allover US and EP patstat publications, by publication year



1978
99,7%
1996
84,0%
1979
97,7%
1997
81,3%
1980
96,8%
1998
81,5%
1981
93,0%
1999
80,0%
1982
90,3%
2000
76,3%
1983
90,3%
2001
68,1%
1984
89,3%
2002
50,6%
1985
87,4%
2003
47,8%
1986
85,6%
2004
44,5%
1987
84,0%
2005
41,2%
1988
82,5%
2006
44,7%
1989
83,2%
2007
42,8%
1990
80,5%
2008
41,8%
1991
78,0%
2009
41,0%
1992
79,1%
2010
45,0%
1993
82,6%
2011
47,2%
1994
85,0%
2012
43,2%
1995
83,1%



Anyway several caveat must be introduced:

1) Number of claims depends on publication language
Same patent, issued in German may have 1 or 2 claims less than in English

2) Number of claims may be different in different publications of same patent

 Take FI patent EP  1186609; in TLS211 for october 2012 you will find

26, 'EP', '        1186609', 'B1', 10, '2008-12-03', 'DE', 1, '0008'
27, 'EP', '        1186609', 'A3', 10, '2002-10-02', 'DE', 0, ''
28, 'EP', '        1186609', 'A2', 10, '2002-03-13', 'DE', 0, '0009'

that means publication kind A2 lists 9 claims, while B1 lists only 8.

And if we check on original documents:

http://worldwide.espacenet.com/publicationDetails/claims?CC=EP&NR=1186609A2&KC=A2&FT=D&ND=3&date=20020313&DB=worldwide.espacenet.com&locale=en_EP


http://worldwide.espacenet.com/publicationDetails/claims?CC=EP&NR=1186609B1&KC=B1&FT=D&ND=4&date=20081203&DB=worldwide.espacenet.com&locale=en_EP

We see in phase B1 one claim has been withdrawn.
Now it depends from our research question to decide if it is better to consider older (A2) or more recent (B1) number of claims.