more on Claiming a Threat to Innovation, Group Seeks to Overturn 10 Patents

[David Farber <dave () farber net>]

Begin forwarded message:

From: Gregory Aharonian <srctran () world std com>
Date: July 5, 2004 1:52:43 PM EDT
To: dave () farber net
Subject: Re: [IP] Claiming a Threat to Innovation, Group Seeks to 
Overturn 10 Patents

Dave,
        Attached is some software patent statistics, documenting the
continual worsening handling of software patents by the Patent Office
(which is why I have been fighting for ten years, sadly with little
success, to get the Patent Office to improve their logistics).  By
the end of 2004, there should be about 200,000 software patents, which
makes the EFF effort to attack ten patents a silly waste of money.
Every week law firms and companies are attacking dozens of software
patents.
        Feel free to share this data with IP.

Greg Aharonian
Internet Patent News Service


SOFTWARE PATENT STATISTICS 1976-2003

              IN THE LAST TEN YEARS, QUALITY HAS CONTINUED TO DROP

                            Gregory Aharonian, Editor
                     Internet Patent News Service, July 2004
                        www.bustpatents.com/softpats.htm

    I have crunched the data and updated my software patent statistics
to be current as of 2003 (a similar study will be done for nanotech).
I focus your attention on two aspects - worsening pendency times and
worsening amounts of non-patent prior art citations - together well
correlated with decreasing software patent quality for the last ten 
years.
By the end of 2004, there should be over 200,000 software patents.

    In terms of pendency, software patents for many years have taken 
about
five to six more months to prosecute than electronics patents in 
general.
Pendency has risen about two months per year for the last six years,
meaning that in recent years PTO management has failed to manage to have
sufficient numbers of examiners available, let alone do anything about
patent quality.

    Also for at least the last six years, while examiners have about ten
percent more claims to process for software patents as compared to
electrical patents, it is taking examiners about twenty percent more 
time
to process the software patents.  So despite all the PTO pablum to the
contrary, for at least the last six years, PTO management has made 
negative
progress in improving software patent quality - they have completely 
failed.
The pathetic aid provided by the software industry is just that - 
pathetic.

    In terms of patent quality, as measured by non-patent prior art
reference statistics (and non-patent prior art is the best materials to
use to invalidate most software patents), software patent quality has 
not
changed much in over 20 years.  Despite the ever increasing volume of
non-patent prior art, the average software patent is citing only 1 to 2
non-patent prior art items - that is ludicrous.  Second, the vast 
majority
of software patents (about 60%) still cite no non-patent prior art,
indicating that software patent seekers still refuse to take quality
seriously, and that the PTO has failed completely in providing examiners
with the time and resources to search out the non-patent prior art - a
failure that has lasted for decades with the apparent blessing of the
software industry and Congress.

    Think of it this way - there are at least ten times more software
journal articles and conference papers than there are software patents.
If the average software patent is citing about one dozen previous 
software
patents, then the average software patent should be citing about one
hundred non-patent prior art items to be a statistically similar 
sampling
of relevant prior art.  But the average software patent is only citing
about two non-patent prior art items, which means that for the most part
issued software patents do no reflect any of the non-patent prior art.
Which means that for the most part, most software patents should be
presumed to be invalid.  Of course, this is true for most electronic
patents, since they also cite paltry amounts of non-patent prior art.
Software patent quality, though, is worse (especially since in the data
below for Electrical patents, software patent data is included - 
separate
out the software patents from the electrical patents, and software 
patent
quality and pendency is even worse).

    One thing to keep in mind - it is reprehensible for the PTO, and the
FTC and NAS studies, to not be providing similar data to the public.  
They
don't do so, because in the end, they just don't care about patent 
quality.
And neither does industry nor the patent bar, which could easily fund a
collection of such data.  Heck, even for the boring/trivial statistics
the PTO does provide the public (like patents awarded to Alaska), most
of its data is current only until to 2001 - HEY it's 2004.  How 
incompetent
and how telling of a lack of interest in quality.

    In the table below, the "Soft Pats" column is the number of software
patents issued in any one year.  The "Numb UREF" column is the number of
prior art references to U.S. patents.  The "Numb OREF" column is the
number of prior art references to non-patent prior art.  The "Perc ZREF"
column is the percentage of patents citing no non-patent prior art. The
"Pend SOFT" is the pendency in months for software patents. The "Pend 
ELEC"
is the pendency in months for all electronic patents (including the
software patents).  The "Claims SOFT" column is the average number of
claims in software patents, while the "CLAIMS ELEC" column is the 
average
number of claims in all electrical patents. The "Perc. Corp." column if
the number of software patents going to corporations.

    All of these statistics rely on a few simplfying assumptions, but 
none
that change the general message of the data and the relative 
comparisons.
NAS or FTC want to share some of their monies so I can do a more 
thorough
analysis?  I think not.




           Soft   Numb   Numb   Perc   Pend   Pend  Claims Claims  Perc.
Year       Pats   UREF   OREF   ZREF   SOFT   ELEC   SOFT   ELEC   Corp.
----      -----   ----   ----   ----   ----   ----  ------ ------  -----
2003      19638   14.1    1.9    58    39.6   32.5   21.4   19.5   95.8
2002      17396   13.5    1.8    59    37.5   30.8   21.1   18.8   95.9
2001      17070   13.5    1.6    61    35.2   29.8   20.3   18.4   95.7
2000      15743   12.4    1.4    65    33.3   28.9   20.3   18.0   95.5
1999      16581   11.4    1.3    66    31.1   27.4   19.3   17.1   95.4
1998      15606   11.4    1.3    66    29.8   26.4   18.4   16.6   95.5
1997       9250   10.9    1.2    63    29.0   25.4   17.4   15.8   95.0
1996       8859   10.3    1.2    63    26.8   24.0   16.3   15.3   95.2
1995       6789    9.5    1.1    64    26.2   23.0   15.3   14.6   94.6
1994       6117    8.8    1.0    66    27.6   23.4   14.9   14.2   94.8
1993       4938    8.6    0.9    67    28.5   23.4   14.7   13.8   94.8
1992       3846    9.1    0.8    69    28.3   22.3   14.2   13.7   94.5
1991       3624    8.1    0.8    72    26.1   21.6   13.6   13.2   94.1
1990       3373    8.3    0.7    73    25.7   21.2   13.7   12.9   94.3
1989       3780    8.3    0.8    72    27.3   22.0   13.7   12.6   93.7
1988       2646    7.4    0.6    74    29.3   23.2   12.1   11.7   93.7
1987       2463    7.4    0.5    78    30.9   25.3   12.4   11.4   93.8
1986       2030    7.5    0.5    79    32.6   27.7   12.3   11.1   93.3
1985       1915    7.5    0.4    80    32.1   28.2   12.4   11.0   93.1
1984       1636    7.3    0.4    83    32.6   28.6   11.9   10.7   92.0
1983       1234    7.0    0.3    85    32.5   28.6   12.0   10.6   92.7
1982       1206    7.0    0.3    85    30.0   27.3   12.7   10.4   89.1
1981       1102    7.1    0.2    88    27.3   25.9   12.4   10.5   90.2
1980       1050    6.8    0.2    91    27.2   26.9   12.6   10.6   90.6
1979        749    6.3    0.1    94    23.6   22.7   11.7   10.5   90.1
1978        986    5.9    0.0    97    21.5   21.4   12.7   10.4   90.2
1977        856    5.5    0.1    97    21.1   21.0   12.1   10.1   89.6
1976        657    5.1    0.1    96    21.4   22.1   12.0    9.9   86.3
         ------
         171140

-------------------------------------
You are subscribed as interesting-people () lists elistx com
To manage your subscription, go to
 http://v2.listbox.com/member/?listname=ip

Archives at: http://www.interesting-people.org/archives/interesting-people/