[Corpora-List] Patent application for "Referent tracking of portions of reality"

Wed Sep 1 19:24:17 UTC 2010

Patent infringement is one thing that I vaguely worry about. Very vaguely. I have only had one idea that I considered worth patenting, but didn't. I think that as academics we would naturally look to trace the intellectual history of our ideas before applying for a patent. 

I have also noticed clear cases of independently innovations -- the use of font size or line thickness to represent frequency or collocational strength and the use of colour coding to indicate POS information. It would be upsetting if someone patented this type of rather natural development, or the KWIC format for that matter.

Generally, I take comfort from the fact that corpus tools are rather small fry in the general order of things and are not either mass market (like social networking) or aimed at enterprises with deep pockets (like text analytics).  This is taking a narrow view, however, and I would like to see a much more stringent process for gaining patents.

Michael
---
Michael Barlow
Assoc. Prof. DALSL. University of Auckland
www.michaelbarlow.com
________________________________________
From: corpora-bounces at uib.no [corpora-bounces at uib.no] On Behalf Of Mike Scott [mike at lexically.net]
Sent: Thursday, September 02, 2010 5:57 AM
To: corpora at uib.no
Subject: Re: [Corpora-List] Patent application for "Referent tracking of portions of reality"

  WordSmith Tools has been referred to a couple of times. Speaking for
myself I do not think I have created anything which I could imagine
myself (ethically) attempting to patent. The kinds of commonly-processed
lists such as concordances, key word plots etc. WordSmith creates are
open to anyone to reproduce -- AntConc and WMatrix have both done so,
with my expressed enthusiasm behind them -- and I didn't invent the idea
of a concordance or key words either, just implemented them.

I do agree there is always a danger of somebody causing a painful and
expensive nuisance in the ways discussed, however. There is also some
danger of getting food poisoning or having a motorbike crash but I still
ride a bike and eat.

Cheers -- Mike

On 01/09/2010 17:43, Laurence Anthony wrote:
> Hi again,
>
> This discussion of patents is an important discussion that more of us
> in the world of corpus linguistics (which in some way is part of
> software engineering) should be concerned about.
>
>> I would like to reiterate the point of my previous note:  the so-called
>> "inventors" have been patenting methods that were in the public domain
>> for years.  Publication is the most effective way to protect oneself
>> (and the whole community) against lawsuits.
>>
>> But if patents of public domain methods are granted (as they have
>> been in one of the patents, and as the inventors threaten to do
>> in the other patent application), then anybody who uses those
>> public domain methods is in danger of being sued.
> I still think the comment "anybody who uses those public domain
> methods is in danger of being sued." is a little too strong. Are there
> any patent lawyers in the discussion list that can give a more
> definitive answer about the potential risks here?
>
>> JFS:
>>>> But you can change the names of all the variables and rewrite all
>>>> the comments in a program, and every variation will compute exactly
>>>> the same results.  If somebody is granted a software patent, they
>>>> can threaten to sue anybody who uses an algorithm that is provably
>>>> equivalent.
>> LA:
>>> This is not correct, at least in Japan where I have successfully
>>> applied for and received two patents (except possibly the last line).
>> Both sentences are true.  If you relabel all the names, you still
>> have exactly the same algorithm.
>>
> I may have misunderstood your  original statement, but it seemed that
> the implication was that a case could be made against someone
> producing the same *results*. I would still say that this is
> incorrect. However, I think you and I agree that a case can be made
> against software that uses an identical *algorithm*.
>
>> LA:
>>> This also seems incorrect, in Japan at least. If you publish something
>>> (in a journal for example), it automatically becomes public domain.
>> JFS:
>> By "protection", I meant protection against being sued.  I have been
>> publishing books and papers about conceptual graphs for 34 years, and
>> I would prefer not to be sued for using those methods.
>> LA:
>>> If someone does manage to get a patent on their software, we still
>>> have nothing to worry about because we are using and developing
>>> software based on much earlier public-domain ideas.
> JFS:
>> There is something to worry about:  the patent may be unenforceable
>> if carefully examined in court.  But the people who hold the patent
>> can still threaten to sue, and defending against such a lawsuit can
>> be very expensive.
> I see what your argument is now. You are worried that some patent
> holder could sue someone else (even if/when they have no chance of
> winning). And, that person would then have the grief of trying to
> defend their perfectly legal methods/tools in an expensive court case.
>
> So, coming back to an earlier comment, you are worried that the day
> might come when Mike Scott could get sued for developing WordSmith
> Tools, because some person has got a patent on this public domain
> idea. It's becoming a very sad world when we have to think about this
> kind of problem, isn't it?!
>
> I think this might be a greater concern for those who live in the
> states, where it is more common to sue people (at least that's what we
> hear). Here in Japan, such cases very, very rarely get to court. In
> the end, I think it's incredibly difficult to defend ourselves against
> trouble of this sort. And, an escalation in protective measures just
> leads to an increasingly worse situation.
>
> So, assuming you think this is a serious problem, what would you say
> the corpus community should do from now?
>
> Laurence.
>
> _______________________________________________
> Corpora mailing list
> Corpora at uib.no
> http://mailman.uib.no/listinfo/corpora

--
Mike Scott

***
If you publish research which uses WordSmith, do let me know so I can include it at
http://www.lexically.net/wordsmith/corpus_linguistics_links/papers_using_wordsmith.htm
***

Aston University and Lexical Analysis Software Ltd.
Room NX01
Aston University
Aston Triangle, Birmingham, B4 7ET
0121 204 3809

www.lexically.net

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora