RE: [PIUG List] Beyond Keywords: Concept searching for patents

Dear Bob,

I was delighted to see your nice comment on the EI EnCompass
Databases(formely API EnCompass Databases). Our thesaurus is helping
searchers find relevant items through use of controlled vocabulary.

Kind regards,

Perry Nowacki
Technical Information Manager
Engineering Information
360 Park Avenue South
New York, NY 10010-1710 USA
Tel +1 212 633 3863
Fax +1 212 633 3650

-----Original Message-----
From: BOB BUNTROCK [mailto:buntrock16_at_verizon.net]
Sent: Monday, April 21, 2008 10:50 PM
To: PIUG Discussion List @ Listbox
Subject: Re: [PIUG List] Beyond Keywords: Concept searching for patents

(At the risk of receiving >40 out of office messages, I'll try answering
this)

This searching method may be superior to classification searching but
may not be superior to good ("old fashioned"?) indexing. This latter
method--good abstracting and indexing--is why, for chemical patents at
least, Chemical Abstracts and Derwent were so successful in serving the
searching needs of patent searchers the world over. Even better,
construct and use a good thesaurus as did API (American Petroleum
Institute) and its users for decades.

A not so great comparison can be made to finding a book in the library.
Classification can get you into one or more promising niches (but may
miss
some) but then you must do some browsing within that classification to
find other material. Searching wise, the best methods use both
classification, indexing, and yes, even full text searching to answers
questions and solve problems.

Your method sounds an attempt at some form of automatic indexing,
clustering, whatever, and is probably superior to a Google-like search
engine, but that may not be saying much. I may be in the old fogey
class, but I'll trust a good A&I producer much more than non-human
parsers, 1st order search engines (does Goggle come anywhere near
meeting Salton's methods?), or whatever.

Proof's in the pudding. Run some comparisons with realistic search
questions on your system, a number of full text search engines,
classification databases, and appropriate A&I databases. We'll look
forward to the results.

-- Bob Buntrock
(veteran of the wars for longer than I care to admit) Orono, ME

----- Original Message -----
From: "James Ryley" <jryley_at_freepatentsonline.com>
To: "PIUG Discussion List @ Listbox"
<piug_discussion_list_at_v2.listbox.com>
Sent: Monday, April 21, 2008 4:39 PM
Subject: RE: [PIUG List] Beyond Keywords: Concept searching for patents


> Hi Jeremy,
>
> You don't specify how your technology works, but this sounds similar
to
> LSI
> and clustering. If you are using something based around singular
value
> decomposition and/or some variation of K-means clustering, we, and
several
> other companies are already working on this or have solutions
commercially
> available. If you are working on something more novel, perhaps this
is a
> new tool that could benefit the IP community -- I'd be happy to talk
to
> you
> about it.
>
> Sincerely,
> James Ryley, Ph.D.
> www.SumoBrain.com / www.FreePatentsOnline.com
>
> This communication is to be treated as confidential and the
information in
> it may not be used or disclosed except for the purpose for which it
has
> been
> sent. Nothing contained herein nor on a related web site should be
> construed
> as legal or patenting advice.
>
>
>> -----Original Message-----
>> From: Mail [mailto:jeremy.trash_at_gmail.com]
>> Sent: Monday, April 21, 2008 3:35 PM
>> To: PIUG Discussion List @ Listbox
>> Subject: [PIUG List] Beyond Keywords: Concept searching for patents
>>
>> Please comment on whether you think this is useful, who might
>> want it, or why it's a waste of time.
>>
>> A friend and I researching a new way to search for patents at UT
>> Austin. We call the new method "Concept Searching."
>>
>> Currently people are forced to search on terms that are tied to
form.
>> By the nature of the English language, it is very easy to create
many
>> different descriptions for products that perform the same essential
>> functions. Our algorithm can parse plain text into (partial, but
>> useful) conceptual models as well as the implicit customer needs
from
>> the patent description. This enables two or more products that may
>> appear to be different based on architecture or layout, to be
>> functionally compared to determine true novelty or uniqueness.
>> Conceptual models are universal, machine readable, and have a
>> one-to-one correspondence with products.
>>
>> Thanks!
>>
>> Jeremy Guillory
>> jeremy.guillory_at_gmail.com
>>
>> -------------------------------------------
>> -------------------------------------------
>> PIUG R -- Celebrating Our 20th Anniversary
>>
>> PIUG Discussion List @ Listbox
>> List owner: Patent Information Users' Group Inc., http://www.piug.org
>> List rules and guidelines: http://www.piug.org/list.php
>> List admins: piug_discussion_list-owner(at)v2(dot)listbox(dot)com
>>
>> Modify Your Subscription:
>>
> http://www.listbox.com/member/?&
>> Unsubscribe Now:
>> d1949b
>
> -------------------------------------------
> -------------------------------------------
> PIUG (r) -- Celebrating Our 20th Anniversary
>
> PIUG Discussion List @ Listbox
> List owner: Patent Information Users' Group Inc., http://www.piug.org
> List rules and guidelines: http://www.piug.org/list.php
> List admins: piug_discussion_list-owner(at)v2(dot)listbox(dot)com
>
> Modify Your Subscription:
> http://www.listbox.com/member/?&
> Unsubscribe Now:
>


-------------------------------------------
-------------------------------------------
PIUG (r) -- Celebrating Our 20th Anniversary

PIUG Discussion List @ Listbox
List owner: Patent Information Users' Group Inc., http://www.piug.org
List rules and guidelines: http://www.piug.org/list.php
List admins: piug_discussion_list-owner(at)v2(dot)listbox(dot)com

Modify Your Subscription:
http://www.listbox.com/member/?&
606
Unsubscribe Now:
7-1db678

-------------------------------------------
-------------------------------------------
PIUG ® -- Celebrating Our 20th Anniversary

PIUG Discussion List @ Listbox
List owner: Patent Information Users' Group Inc., http://www.piug.org
List rules and guidelines: http://www.piug.org/list.php
List admins: piug_discussion_list-owner(at)v2(dot)listbox(dot)com

3&id_secret=101952990-a5e9bf
_secret=101483107-69b265

Received on Tue Apr 22 2008 - 15:12:49