<meta http-equiv="Content-Type" content="text/plain; charset=utf-8">

#HtW: How it was done

#OeW: Top EPO Software Patent Applicants 1978-2003

#btY: Publication Years

#ptt: Applicant Countries

#invland: Inventor Countries

#Stt: Old Statistics

#Tke: The difficulty is that there is no explicit identification for
software related patents. Therefore we had to download and scan all
patents, so as to assign each patent a probability of being a software
patent.  Expressions related to algorithms and programming raised the
probability, those related to forces of nature lowered it.  This has
led to a fairly high score rate, but still you will occasionally
encounter a true hardware patent.

#aso: ../txt/ep/1276/061

#Cyi: Country Statistics

#Cyi2: Company Statistics

#tiW: a database of patent statistics (based on a bibliometric search
profile) concerning computer software based patent application in
European countries, Japan and USA (22 countries in all). This can be
used to trace the development of embedded software in different
countries, various patent classes and groups, etc, as a function of
time for the past fourteen years. ...

# Local Variables: ;
# coding: utf-8 ;
# srcfile: /usr/share/emacs/site-lisp/phm/app/swpat/swpatpikta.el ;
# mailto: mlhtimport@ffii.org ;
# login: phm ;
# passwd: YYYYY ;
# feature: swpatdir ;
# dok: swpiknamcu ;
# txtlang: en ;
# multlin: t ;
# End: ;