---

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Linda Farmer
Sent: Friday, April 07, 2006 10:28 AM
To: Tax CoP
Subject: [TaxoCoP] Auto-categorization tools

Hi Becky,

Here is a list of auto-categorization tools which usually incorporate various degrees of taxonomy management. Some CMS vendors fold in auto-categorization functionality into their products (e.g. Stellent) but are not included here. This list includes only those tools which offer their auto-categorization/taxonomy management products as stand-alone or linkable to other applications through APIs. Schemalogic is not an auto-categorization tool but provides very powerful and comprehensive metadata and taxonomy management that serves up taxonomies and metadata to applications like auto-categorization and CMS.

Could well be I've missed some auto-categorization/taxonomy tool out there. Those in the know, please let me know so we can build a complete list. Perhaps the list should be broadened to include other applications that incorporate significant auto-categorization and taxonomy management features into their product mix e.g. IBM, Inmagic products. Note: Stratify (http://www.stratify.com) is a major vendor of this type of tool but is now focusing on the legal domain.

Auto-Categorization Tools

Autonomy: http://www.autonomy.com (Includes Verity)

Convera: http://www.convera.com

Data Harmony: http://www.dataharmony.com

Endeca: http://endeca.com

Entrieva: http://www.entrieva.com

Interwoven: http://www.interwoven.com

Inxight: http://www.inxight.com

Nstein Technologies: http://www.nstein.com

SchemaLogic: http://schemalogic.com

Siderean: http://www.siderean.com

Smartlogik: http://www.aprsmartlogik.com

Teragram: http://www.teragram.com

Vivisimo: http://vivisimo.com

Wordmap: http://www.wordmap.com

Regards,

Linda Farmer

Second Knowledge Solutions

http://k2s.ca

email: lfarmer@...

voice: 905-465-3387

On 6-Apr-06, at 12:36 PM, Becky Sherman wrote:

I wanted to know what products or tools do you currently have that help you manage your taxonomy? I heard someone talking about Data Harmony and wanted to know what other companies do this type of stuff - as I am trying to learn about what is out there because we don't have a really knowledge management area or tools in the enterprise - so data is everywhere. All help is good help - b

---

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Farida Hasanali
Sent: Friday, April 07, 2006 12:34 PM
To: TaxoCoP@yahoogroups.com
Subject: RE: [TaxoCoP] Auto-categorization tools

Hello folks,

I just came across this page as I was searching the Library of Congress and it has links to several resources.  Thought since there seems to be some discussions of tools going on this might be appropriate.

http://www.loc.gov/flicc/wg/taxonomy.html

Also Seth, there is a section for Groups/Discussions with links under it.  You might want to contact the author and see how we can get the link to our group on there too.

Take care,

farida

Seth Earley <seth@...> wrote:

Does Synaptica do auto categorization?  I am familiar with it’s powerful taxonomy and thesaurus management capabilities but was not aware of auto cat functions.

 

Seth

 

Subject: RE: [TaxoCoP] Auto-categorization tools

 

Hi Linda,

 

You may also wish to include Synaptica: http://www.factiva.com/products/taxonomy/synaptica.asp?node=menuElem1511 on your list.

 

Regards

 

Chris

 

---

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Linda Farmer
Sent: 07 April 2006 15:28
To: Tax CoP
Subject: [TaxoCoP] Auto-categorization tools

Hi Becky,

 

Here is a list of auto-categorization tools which usually incorporate various  degrees of taxonomy management.  Some CMS vendors fold in  auto-categorization functionality into their products (e.g. Stellent) but are not included here. This list includes only those tools which offer their auto-categorization/taxonomy management products as stand-alone or linkable to other applications through APIs.  Schemalogic is not an auto-categorization tool but provides very powerful and comprehensive metadata and taxonomy management that serves up taxonomies and metadata to applications like auto-categorization and CMS. 

 

Could well be I've missed some auto-categorization/taxonomy tool out there. Those in the know, please let me know so we can build a complete list. Perhaps the list should be broadened to include other applications that incorporate significant auto-categorization and taxonomy management features into their product mix e.g. IBM, Inmagic products. Note: Stratify (http://www.stratify.com) is a major vendor of this type of tool but is now focusing on the legal domain. 

 

 

Auto-Categorization Tools

 

Autonomy: http://www.autonomy.com (Includes Verity)

Convera: http://www.convera.com

Data Harmony: http://www.dataharmony.com

Endeca: http://endeca.com

Entrieva: http://www.entrieva.com

Interwoven: http://www.interwoven.com

Inxight: http://www.inxight.com

Nstein Technologies: http://www.nstein.com

SchemaLogic: http://schemalogic.com

Siderean: http://www.siderean.com

Smartlogik: http://www.aprsmartlogik.com

Teragram: http://www.teragram.com

Vivisimo: http://vivisimo.com

Wordmap:  http://www.wordmap.com

•  
•  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

 

---

---

How low will we go? Check out Yahoo! Messenger’s low PC-to-Phone call rates.

---

Yahoo! Messenger with Voice. PC-to-Phone calls for ridiculously low rates.

----- Original Message -----

From: Seth Earley

To: TaxoCoP@yahoogroups.com

Sent: Friday, April 07, 2006 12:23 PM

Subject: RE: [TaxoCoP] Auto-categorization tools

Does Synaptica do auto categorization?  I am familiar with it’s powerful taxonomy and thesaurus management capabilities but was not aware of auto cat functions.

 

Seth

 

All - I've been following the threads recently about
auto-categorization tools with great interest.

It appears that most of the interest so far has been in tools to
automatically classify content against a given taxonomy. I've recently
been struggling with what I think is a much simpler, but related
problem.

On our intranet, we use a search engine that produces weekly reports
of the terms that are used.  Like many search engines, the reporting
is based on the exact terms used by the users - right down to the case
of the terms (though the engine itself is case insensitive).

What I've found is that even the most common searches are a
vanishingly small percentage of the overall set of search terms - the
most common search term is commonly around 1% of search terms and the
top 100 terms usually are about 20% of all searches within a week.

Given this, it's really very challenging to see what people are
*really* looking for.  I know that the top search might consistently
be "travel" comprising about 1% of the searches, but I don't know that
there might be another 5% of searches that if I could tell that they
were all related to each other (but just variations) that are also
searches for travel information.

I've taken a stab at reviewing the search terms and doing some manual
categorization to start to "see the forest for the trees", but it's
tedious and time-consuming and, especially when you get another bunch
every week, hard to keep up with.

My question:  Are there any tools that one could apply to the
individual search terms that users are using that would categorize
those into a known taxonomy?

Sorry for the long intro to the question, but I thought I'd provide a
bit of background about why I'm looking for this.

Thanks!
Lee

---

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Lee Romero
Sent: Friday, April 07, 2006 6:05 PM
To: taxocop@yahoogroups.com
Subject: [TaxoCoP] Auto-categorization of search terms?

All - I've been following the threads recently about
auto-categorization tools with great interest.

It appears that most of the interest so far has been in tools to
automatically classify content against a given taxonomy. I've recently
been struggling with what I think is a much simpler, but related
problem.

On our intranet, we use a search engine that produces weekly reports
of the terms that are used.  Like many search engines, the reporting
is based on the exact terms used by the users - right down to the case
of the terms (though the engine itself is case insensitive).

What I've found is that even the most common searches are a
vanishingly small percentage of the overall set of search terms - the
most common search term is commonly around 1% of search terms and the
top 100 terms usually are about 20% of all searches within a week.

Given this, it's really very challenging to see what people are
*really* looking for.  I know that the top search might consistently
be "travel" comprising about 1% of the searches, but I don't know that
there might be another 5% of searches that if I could tell that they
were all related to each other (but just variations) that are also
searches for travel information.

I've taken a stab at reviewing the search terms and doing some manual
categorization to start to "see the forest for the trees", but it's
tedious and time-consuming and, especially when you get another bunch
every week, hard to keep up with.

My question:  Are there any tools that one could apply to the
individual search terms that users are using that would categorize
those into a known taxonomy?

Sorry for the long intro to the question, but I thought I'd provide a
bit of background about why I'm looking for this.

Thanks!
Lee

We looked at Endeca in depth. Really cool tool. It does auto-categorization. It can derive structure where none exist and also utilizes existing metadata.

Omnifind (IBM), Fast Search & Transfer, SPSS (Lexiquest), SER, SAS text miner are others that come to mind.

Interesting you put Vivisimo in the same category. They do not categorize into an existing structure but cluster information on the fly based on the text and metadata received in a search result. All other tools is see in the category entity extraction and/or categorization in the sense of text-taxonomy association.

Karin 

SMS schreiben mit WEB.DE FreeMail - einfach, schnell und    kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192

Back in the year of 2000, I saw demos of auto and
semi_auto categorization by portal vendors such as
Hummingbird, Plumtree, and by intelligent text
processing vendors such as TextWise.

If you are interested in developing taxonomy
applications in the medical field, try Apelon,
formerly called Lexical Technology,and JARG.com.

In addition, a good starting point for finding players
in the IR and IE field is to go to TREC.NIST.GOV.

These are just a few vendors coming to my mind at the
moment.


Amanda Xu


--- Karin Schneider <kschnei1@...> wrote:

>
>
>
> We looked at Endeca in depth. Really cool tool. It
> does auto-categorization. It can derive structure
> where none exist and also utilizes existing
> metadata.
>
> Omnifind (IBM), Fast Search & Transfer, SPSS
> (Lexiquest), SER, SAS text miner are others that
> come to mind.
>
> Interesting you put Vivisimo in the same category.
> They do not categorize into an existing structure
> but cluster information on the fly based on the text
> and metadata received in a search result. All other
> tools is see in the category entity extraction
> and/or categorization in the sense of text-taxonomy
> association.
>
> Karin
>
> *Von:* TaxoCoP@yahoogroups.com
> *Gesendet:* 07.04.06 19:16:08
> *An:* <TAXOCOP@YAHOOGROUPS.COM>
> *Betreff:* RE: [TaxoCoP] Auto-categorization tools
>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?;color:black'>Does Endeca
> doauto-categorization? I thought they were just
> search optimization software. Fromwhat I understood,
> they need pre-existing metadata to create the
> GuidedNavigation.
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?;color:black'>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?;color:purple'>barbaramcglamery
> <spanstyle='font-size:10.0pt;font-family:"Wingdings
> 2?;color:purple'>?lt;FONTSIZE=2 face="Courier New"
> color="purple"> sr. librarian, ontology development
> <FONTSIZE=2 face="Wingdings 2" color="purple">?>
<spanstyle='font-size:10.0pt;font-family:"Courier
> New?;color:gray'>time inc.interactive
> <spanstyle='font-size:10.0pt;font-family:"Wingdings
> 2?;color:purple'>?lt;FONTSIZE=2 face="Courier New"
> color="purple"> 212.522.9198
> ?lt;spanstyle='font-size:10.0pt;font-family:"Courier
> New?;color:purple'> <FONTSIZE=2 face="Courier
> New" color="gray">im:babsmcglamery
>
>
<spanstyle='font-size:10.0pt;font-family:Tahoma'>-----Original
> Message-----
> *From:*
>
TaxoCoP@yahoogroups.com[mailto:TaxoCoP@yahoogroups.com]
> *On Behalf Of*Linda Farmer
> *Sent:* Friday, April 07, 2006 10:28AM
> *To:* Tax CoP
> *Subject:* [TaxoCoP]Auto-categorization tools
>
> <spanstyle='font-size:12.0pt'>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Hi Becky,
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Here is a list ofauto-categorization tools
> which usually incorporate various degrees oftaxonomy
> management. Some CMS vendors fold in
> auto-categorizationfunctionality into their products
> (e.g. Stellent) but are not included here.This list
> includes only those tools which offer
> theirauto-categorization/taxonomy management
> products as stand-alone or linkable toother
> applications through APIs. Schemalogic is not
> anauto-categorization tool but provides very
> powerful and comprehensive metadataand taxonomy
> management that serves up taxonomies and metadata to
> applicationslike auto-categorization and CMS.
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Could well be I've missedsome
> auto-categorization/taxonomy tool out there. Those
> in the know, please letme know so we can build a
> complete list. Perhaps the list should be
> broadenedto include other applications that
> incorporate significant auto-categorizationand
> taxonomy management features into their product mix
> e.g. IBM, Inmagicproducts. Note: Stratify
> (<FONTCOLOR="#001FDD">http://www.stratify.com
> [http://www.stratify.com/])is a major vendor of this
> type of tool but is now focusing on the legaldomain.
>
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <FONTSIZE=1 face="Helvetica">
>
> *<FONTSIZE=2 face="Courier New">Auto-Categorization
> Tools*
>
> <FONTSIZE=1 face="Helvetica">
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Autonomy: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
>
www.autonomy.com>*<spanstyle='color:#001FDD;font-weight:bold'>http://www.autonom\
y.com**<FONTCOLOR="#001C84">
> (Includes Verity)*
>
> *<FONTSIZE=2 face="Courier
>
New">Convera*<spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>: <AHREF="HTTP: ?
> www.convera.com><spanstyle='font-family:"Times
> Roman?;color:#001FDD' New>http://www.convera.com
>
> *<FONTSIZE=2 face="Courier New">Data
>
Harmony*<spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>: <AHREF="HTTP: ?
>
www.dataharmony.com><spanstyle='color:#001FDD'>http://www.dataharmony.com
>
> *<FONTSIZE=2 face="Courier New">Endeca:
> *<spanstyle='font-size:10.0pt;font-family:"Courier
> New?'><FONTCOLOR="#001FDD">http://endeca.com
> [http://endeca.com/]
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Entrieva: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
>
www.entrieva.com>*<spanstyle='color:#001FDD;font-weight:bold'>http://www.entriev\
a.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Interwoven: *<FONTSIZE=2
> face="Courier New"><AHREF="HTTP: ?
>
www.interwoven.com>*<spanstyle='color:#001FDD;font-weight:bold'>http://www.inter\
woven.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Inxight: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
> www.inxight.com>*http://www.inxight.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Nstein Technologies: *<FONTSIZE=2
> face="Courier New"><AHREF="HTTP: ?
> www.nstein.com>*http://www.nstein.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">SchemaLogic: *<FONTSIZE=2
> face="Courier New"><AHREF="HTTP: ?
> schemalogic.com>*http://schemalogic.com*
>
> *<FONTSIZE=2 face="Courier New">Siderean:
> *<spanstyle='font-size:10.0pt;font-family:"Courier
> New?'><AHREF="HTTP: ?
> www.siderean.com>http://www.siderean.com
>
> *<FONTSIZE=2 face="Courier
>
New">Smartlogik:*<spanstyle='font-size:10.0pt;font-family:"Courier
> New?'> <AHREF="HTTP: ?
>
www.aprsmartlogik.com><spanstyle='color:#001FDD'>http://www.aprsmartlogik.com
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Teragram: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
>
www.teragram.com>*<spanstyle='color:#001FDD;font-weight:bold'>http://www.teragra\
m.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Vivisimo: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
> vivisimo.com>*http://vivisimo.com*
>
> *<FONTSIZE=2 face="Courier New"
> color="#001c84">Wordmap: *<FONTSIZE=2 face="Courier
> New"><AHREF="HTTP: ?
> www.wordmap.com>*http://www.wordmap.com*
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Regards,
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Linda Farmer
>
> <spanstyle='font-size:9.0pt;font-family:Helvetica'>
>
>
> <spanstyle='font-size:10.0pt;font-family:"Courier
> New?'>Second
=== message truncated ===


             Amanda Xu
   58-11 197th St.
Fresh Meadows, NY 11365
718-990-6716 (voice)
axu789@... (email)










__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

I replied to soon. You already answered a lot of my questions in other e-mails. Could you elaborate a bit more on the type of rules you can define and what it takes to build/maintain them (time, resources, skill). Does Thes Master integrate with CMS? Does it have import pipe configured for HL-7, Snomed and Meddra? How does it deal with synchronization of updates?

A demo would really be cool.

Thanks,

Karin

SMS schreiben mit WEB.DE FreeMail - einfach, schnell und    kostenguenstig. Jetzt gleich testen! http://f.web.de/?mc=021192

http://www.grokker.com/

I got results from Grokker Search like this:

https://beta.grokker.com/grokker.html?query=Bhojaraju&Yahoo=true&Wikipedia=true&numResults=250
 
Thanks and Regards

Bhojaraju G
Lead - Knowledge Management
***********************************************************************************************************
KM Cyberary:   http://www.bhojarajug.freeservers.com/cyberary.html  
KM-Forum:       http://groups.yahoo.com/group/KM-Forum/
LinkedIn:         https://www.linkedin.com/in/bhojarajug
Skype Me:        bhojarajug

--- In TaxoCoP@yahoogroups.com, Linda Farmer <lfarmer@...> wrote:
>
> Hi Becky,
>
> Here is a list of auto-categorization tools which usually incorporate
> various degrees of taxonomy management. Some CMS vendors fold in
> auto-categorization functionality into their products (e.g. Stellent)
> but are not included here. This list includes only those tools which
> offer their auto-categorization/taxonomy management products as stand-
> alone or linkable to other applications through APIs. Schemalogic is
> not an auto-categorization tool but provides very powerful and
> comprehensive metadata and taxonomy management that serves up
> taxonomies and metadata to applications like auto-categorization and
> CMS.
>
> Could well be I've missed some auto-categorization/taxonomy tool out
> there. Those in the know, please let me know so we can build a
> complete list. Perhaps the list should be broadened to include other
> applications that incorporate significant auto-categorization and
> taxonomy management features into their product mix e.g. IBM, Inmagic
> products. Note: Stratify (http://www.stratify.com) is a major vendor
> of this type of tool but is now focusing on the legal domain.
>
>
>
> Auto-Categorization Tools
>
>
>
> Autonomy: http://www.autonomy.com (Includes Verity)
>
> Convera: http://www.convera.com
>
> Data Harmony: http://www.dataharmony.com
>
> Endeca: http://endeca.com
>
> Entrieva: http://www.entrieva.com
>
> Interwoven: http://www.interwoven.com
>
> Inxight: http://www.inxight.com
>
> Nstein Technologies: http://www.nstein.com
>
> SchemaLogic: http://schemalogic.com
>
> Siderean: http://www.siderean.com
>
> Smartlogik: http://www.aprsmartlogik.com
>
> Teragram: http://www.teragram.com
>
> Vivisimo: http://vivisimo.com
>
> Wordmap: http://www.wordmap.com
>
>
> Regards,
>
> Linda Farmer
>
> Second Knowledge Solutions
> http://k2s.ca
> email: lfarmer@...
> voice: 905-465-3387
>
>
> On 6-Apr-06, at 12:36 PM, Becky Sherman wrote:
> I wanted to know what products or tools do you currently have that
> help you manage your taxonomy? I heard someone talking about Data
> Harmony and wanted to know what other companies do this type of stuff
> - as I am trying to learn about what is out there because we don't
> have a really knowledge management area or tools in the enterprise -
> so data is everywhere. All help is good help - b

Hi all,

Not to be a wiki-pusher, but I think this topic would lend itself
perfectly to elaboration on the wiki.
All of you who have mentioned tools or have specifications about what
a tool actually does, please visit the wiki and add a comment under
the tool (or add a new one to the list).
http://taxocop.wikispaces.com/TaxoTools

That way we can keep a running tab on all of this valuable content.

Thanks!
Stephanie

You wiki pusher you...  ;-)

I agree.  And Stephanie you have been doing a great job on the wiki.  A
couple of people had offered to create synopses of CoP calls and post them -
I won't name names (yet...).

Perhaps I can also encourage others to stop in and post their profiles on
the wiki. If you have some thoughts on tools - feel free to add them.

(And Kate - now that I am spelling it correctly will you stop in?> <smile>)

Here is the url http://taxocop.wikispaces.com/

Seth
-----------
Subject: [TaxoCoP] Auto-categorizing tools - a perfect topic for the wiki

Hi all,

Not to be a wiki-pusher, but I think this topic would lend itself
perfectly to elaboration on the wiki.
All of you who have mentioned tools or have specifications about what
a tool actually does, please visit the wiki and add a comment under
the tool (or add a new one to the list).
http://taxocop.wikispaces.com/TaxoTools

That way we can keep a running tab on all of this valuable content.

Thanks!
Stephanie

Hello! I have been a lurker for the last year and was recently
reminded by Seth to introduce myself.

My name is Alexandra Proserpio from Sandia National Laboratories. I
support corporate classification efforts for search and
browse (e.g., A-Z Index) applications. I also consult with SME and
developers on IA issues. My educational background is in library
science.

A few years ago, we deployed a 2K node taxonomy (Verity's VIC)
to improve intranet search. It turned out that searchers only
browsed the taxonomy when the quality of search results met their
needs which wasn't often. So we refocused our efforts to improve the
quality of our search engine, developed best bets and a federated
search and began qualitative and quantitative testing. In a recent
search satisfaction survey, 69% of respondents indicated they find
what they are looking for 75 to 100% of the time. We are currently
exploring the value of semantic technologies. I hope to leverage the
taxonomy (with many tweaks) if we decide to go that route. On a
related project, we are also exploring the value and opportunities
for enterprise taxonomy.

A few questions to members:

1) For those of you who lead/manage an enterprise taxonomy, are
you required to prove ROI? If so, how often and how do you do this?

2) In your organizations, do you have a (formal or informal)
taxonomy (or ontology) COP, COI or user groups?

Lastly, I'd like to share an out of print report I came across
recently; it may be of some interest for members. Check with an
academic library in your area.

Gilchrist, Alan and Peter Kibby. Taxonomies for business: access and
connectivity in a wired world. London. TFPL Ltd. 2000.  Here is a
brief snapshot of the report:
http://www.bokis.is/iod2001/papers/Gilchrist_paper.doc

Thanks. I look forward to contributing to the group!

Alexandra M. Proserpio

Greetings!

I've just recently joined the group, though I've been meaning to do so
for quite some time. I am a professor at the University of Illinois in
Champaign-Urbana and teach in the area of information organization. As
a faculty member, I am planning to be involved in taxonomy creation
and maintenance with some of the business partners at our school (in
addition to teaching in this area) so the experiences and opinions of
the members of this group are most informative. I've also been
impressed by the level of discussions here - many thanks to Seth for
hosting the group and getting things started.

I wanted to add to Alexandra's comment about Alan Gilchrist's work and
let you know that he, along with Jean Aitchison and David Bawden (all
three practitioners who have been active in the controlled vocabulary
world)  has long published a useful handbook on thesaurus construction
(most recent edition - 2000). Amazon allows you to look inside the
book here:
http://www.amazon.com/gp/product/0851424465/ref=pd_ecc_rvi_2/103-4940974-4467841\
?%5Fencoding=UTF8

This is an approachable manual for taxonomy development and
maintenance - not a theoretical treatise. I also note that Gilchrist
also has a forthcoming book on the semantics of retrieval due out in
June - I'll be back in touch as I find out more .


Best,

Kathryn

Well Kathryn, it was about time!!!

<smile>

Kathryn was one of the early CoP presenters (not to be confused with the
Earley CoP presenters... ;-) ) who spoke about faceted search.

By the way, I am still waiting for more volunteers to summarize past call
archives.  Come on now, each is a $50 value.  OK, now I will name names -
Connie, Mark - didn't each of you take on an archive to write up?

(Hey I know all are busy - but nothing like peer pressure to get things
going.)

Seth

Seth Earley
Earley & Associates, Inc
781-444-0287
781-820-8080 cell
Next taxo conference call April 26th, 2 PM EST
"Metrics and Measurement"
Registration and agenda at www.earley.com/events.htm

Taxonomy Community of Practice
http://finance.groups.yahoo.com/group/TaxoCoP

Social Network Analysis Community of Practice
http://finance.groups.yahoo.com/group/ona-prac/


-----Original Message-----
From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of
Kathryn La Barre
Sent: Sunday, April 09, 2006 1:12 PM
To: TaxoCoP@yahoogroups.com
Subject: Re: [TaxoCoP] A member introduction, questions and a taxonomy
resource

Greetings!

I've just recently joined the group, though I've been meaning to do so
for quite some time. I am a professor at the University of Illinois in
Champaign-Urbana and teach in the area of information organization. As
a faculty member, I am planning to be involved in taxonomy creation
and maintenance with some of the business partners at our school (in
addition to teaching in this area) so the experiences and opinions of
the members of this group are most informative. I've also been
impressed by the level of discussions here - many thanks to Seth for
hosting the group and getting things started.

I wanted to add to Alexandra's comment about Alan Gilchrist's work and
let you know that he, along with Jean Aitchison and David Bawden (all
three practitioners who have been active in the controlled vocabulary
world)  has long published a useful handbook on thesaurus construction
(most recent edition - 2000). Amazon allows you to look inside the
book here:
http://www.amazon.com/gp/product/0851424465/ref=pd_ecc_rvi_2/103-4940974-446
7841?%5Fencoding=UTF8

This is an approachable manual for taxonomy development and
maintenance - not a theoretical treatise. I also note that Gilchrist
also has a forthcoming book on the semantics of retrieval due out in
June - I'll be back in touch as I find out more .


Best,

Kathryn



Yahoo! Groups Links

Hi Alexandra - Nice to see you surface here...  (Alexandra and I have met on
a number of occasions and each time she has promised to send a note of
introduction...)

Looks like our wait was well worth it - great resources.  Thanks from all of
us.

(This looks like material that could also be well placed on the wiki -
http://taxocop.wikispaces.com/ as a reminder.  (Hey if everyone put the wiki
address in their signatures perhaps that would improve our critical mass
there...)

Alexandra - would you like to present your statistics and thought process on
our next CoP call?  (Anyone else have an approach to ROI and metrics?)

Seth

Seth Earley
Earley & Associates, Inc
781-444-0287
781-820-8080 cell
Next taxo conference call April 26th, 2 PM EST
"Metrics and Measurement"
Registration and agenda at www.earley.com/events.htm

Taxonomy Community of Practice
http://finance.groups.yahoo.com/group/TaxoCoP

Social Network Analysis Community of Practice
http://finance.groups.yahoo.com/group/ona-prac/


-----Original Message-----
From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of
Alexandra
Sent: Sunday, April 09, 2006 11:48 AM
To: TaxoCoP@yahoogroups.com
Subject: [TaxoCoP] A member introduction, questions and a taxonomy resource

Hello! I have been a lurker for the last year and was recently
reminded by Seth to introduce myself.

My name is Alexandra Proserpio from Sandia National Laboratories. I
support corporate classification efforts for search and
browse (e.g., A-Z Index) applications. I also consult with SME and
developers on IA issues. My educational background is in library
science.

A few years ago, we deployed a 2K node taxonomy (Verity's VIC)
to improve intranet search. It turned out that searchers only
browsed the taxonomy when the quality of search results met their
needs which wasn't often. So we refocused our efforts to improve the
quality of our search engine, developed best bets and a federated
search and began qualitative and quantitative testing. In a recent
search satisfaction survey, 69% of respondents indicated they find
what they are looking for 75 to 100% of the time. We are currently
exploring the value of semantic technologies. I hope to leverage the
taxonomy (with many tweaks) if we decide to go that route. On a
related project, we are also exploring the value and opportunities
for enterprise taxonomy.

A few questions to members:

1) For those of you who lead/manage an enterprise taxonomy, are
you required to prove ROI? If so, how often and how do you do this?

2) In your organizations, do you have a (formal or informal)
taxonomy (or ontology) COP, COI or user groups?

Lastly, I'd like to share an out of print report I came across
recently; it may be of some interest for members. Check with an
academic library in your area.

Gilchrist, Alan and Peter Kibby. Taxonomies for business: access and
connectivity in a wired world. London. TFPL Ltd. 2000.  Here is a
brief snapshot of the report:
http://www.bokis.is/iod2001/papers/Gilchrist_paper.doc

Thanks. I look forward to contributing to the group!

Alexandra M. Proserpio

I have used NetOwl back in the late '90 for named
entity extraction and text summarization.  It supports
XML/OWL:

http://www.netowl.com/products/extractor.html



Amanda Xu

--- Seth Earley <seth@...> wrote:

> Not sure if anyone mentioned Teragram.  Their
> algorithms are part of FAST
> and Verity.
>
>
>
> I had also thought of Endeca as primarily to
> leverage existing metadata.
>
>
>
> I would like to do a series of conference calls on
> search technologies.  The
> challenge is getting good case studies without
> having the vendors present
> (invariably the lapse into vendor pitch mode. - even
> when they are given
> strict instructions not to do so.)
>
>
>
> Perhaps we can post something on the wiki  Anyone
> interested in sharing
> their experiences with the tools that have been
> mentioned?
>
>
>
> The topic for the May Taxo con call will be "To
> Google or Not to Google".
> Who has done a Google eval who'd like to share?
>
>
>
> Seth
>
>
>
> Seth Earley
>
> Earley & Associates, Inc
>
> 781-444-0287
>
> 781-820-8080 cell
>
> Next taxo conference call April 26th, 2 PM EST
> "Metrics and Measurement"
>
> Registration and agenda at
> <http://www.earley.com/events.htm>
> www.earley.com/events.htm
>
> Taxonomy Community of Practice
>
>  <http://finance.groups.yahoo.com/group/TaxoCoP>
> http://finance.groups.yahoo.com/group/TaxoCoP
>
>
>
> Social Network Analysis Community of Practice
>
>  <http://finance.groups.yahoo.com/group/ona-prac/>
> http://finance.groups.yahoo.com/group/ona-prac/
>
>
>
>   _____
>
> From: TaxoCoP@yahoogroups.com
> [mailto:TaxoCoP@yahoogroups.com] On Behalf Of
> Karin Schneider
> Sent: Friday, April 07, 2006 9:06 PM
> To: TaxoCoP@yahoogroups.com
> Subject: [SPAM?] RE: [TaxoCoP] Auto-categorization
> tools
>
>
>
>
> We looked at Endeca in depth. Really cool tool. It
> does auto-categorization.
> It can derive structure where none exist and also
> utilizes existing
> metadata.
>
> Omnifind (IBM), Fast Search & Transfer, SPSS
> (Lexiquest), SER, SAS text
> miner are others that come to mind.
>
> Interesting you put Vivisimo in the same category.
> They do not categorize
> into an existing structure but cluster information
> on the fly based on the
> text and metadata received in a search result. All
> other tools is see in the
> category entity extraction and/or categorization in
> the sense of
> text-taxonomy association.
>
> Karin
>
>


             Amanda Xu
   58-11 197th St.
Fresh Meadows, NY 11365
718-990-6716 (voice)
axu789@... (email)










__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

Alex:

When I started my project on the life-cycle analysis
of journals and books in the domain of business
process management a year and a half ago, I realized
that it was impossible to obtain a 360 degree view of
the workflow process analysis for library journals and
books of all types unless we have a solid enterprise
architecture, which seemed to be a journey of Odyssey
for me.

My comment to your 1st question is that one
possibility to prove ROI is to go beyond search and
retrieval, and get into business intelligence.

When BI comes into play, you will be able to answer a
lot of mission critical questions for your
organization in a systematic and robust way if you
have a solid and scalable enterprise taxonomy.

This perspective has great impact on my design of
taxonomy which requires me to design categories for EA
in panorama view rather than one data set or data
store at a time.

Anyhow, I am still learning and doing EA and taxonomy
construction myself.  This list has provided me
valuable information.

Thanks to everyone!

Amanda Xu






--- Alexandra <amprose2001@...> wrote:

> A few questions to members:
>
> 1) For those of you who lead/manage an enterprise
> taxonomy, are
> you required to prove ROI? If so, how often and how
> do you do this?
>
> 2) In your organizations, do you have a (formal or
> informal)
> taxonomy (or ontology) COP, COI or user groups?
>
> Lastly, I'd like to share an out of print report I
> came across
> recently; it may be of some interest for members.
> Check with an
> academic library in your area.
>
> Gilchrist, Alan and Peter Kibby. Taxonomies for
> business: access and
> connectivity in a wired world. London. TFPL Ltd.
> 2000.  Here is a
> brief snapshot of the report:
>
http://www.bokis.is/iod2001/papers/Gilchrist_paper.doc
>
> Thanks. I look forward to contributing to the group!
>
> Alexandra M. Proserpio
>
>
>
>
>


             Amanda Xu
   58-11 197th St.
Fresh Meadows, NY 11365
718-990-6716 (voice)
axu789@... (email)










__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

Lee -

We have the same "very skinny" Long Tail on our Intranet.  There are
products that "learn" from the user's queries and SERP selections-
Recommendation Engines. Amazon has probably the most famous one of all -
"people who bough xxx also bough yyy."

We have one from Verity. (This Google search highlights some of those
available:
<http://www.google.com/search?q=recommendation+engine&start=0&ie=utf-8&oe=utf-8&\
client=firefox-a&rls=org.mozilla:en-US:official>)

Basically, it grabs data from the clickstream - the user's query, the
results they opened, and the ones they "stayed with." If they use the
browsable taxonomy as part of their search, it also learns which
categories they used. It then can begin making recommendations as part of
your UI. It can recommend any object-type that you have defined: document,
category, person, etc.  It works, but the catch is you have to let the
tool learn behind-the-scenes for a  while before the quality of
recommendations is something you'd be willing to put into production. On
large sites, that can take months. In fact, we didn't care for the results
after 3 months, and then lost a team member and forgot about it. It's not
in our production UI, but the recommendations on the Dev box are getting
better every day!  :)  Another thing for the old To Do list...

I feel a rant coming on about some other aspects of this thread... but
I'll save that for another post!

Cheers!
Christine



>On our intranet, we use a search engine that produces weekly reports
>of the terms that are used.  Like many search engines, the reporting
>is based on the exact terms used by the users - right down to the case
>of the terms (though the engine itself is case insensitive).

>What I've found is that even the most common searches are a
v>anishingly small percentage of the overall set of search terms - the
>most common search term is commonly around 1% of search terms and the
>top 100 terms usually are about 20% of all searches within a week.

>I've taken a stab at reviewing the search terms and doing some manual
>categorization to start to "see the forest for the trees", but it's
>tedious and time-consuming and, especially when you get another bunch
>every week, hard to keep up with.







Christine JM Connors
Metadata Architect

Raytheon Company
Information Technology
600 Technology Park Drive
M/S RTN-2S119
P.O. Box 7038
Billerica, MA 01821
978.436.3195 voice
978.436.3205 fax
508.314.2824 cell Christine_Connors@...

I will add this one to the mix. We use Mercado for our on site search
and they have tool the  they say will do auto classification called
"uniclass". I haven't seen how it works yet but Mercado competes heavily
with Endeca and Fast so I wanted to make sure it was on the list. Thanks
for the feedback on auto-classification this a great! - b


Becky Sherman | www.jcpenney.com |Usability Manager |PH: 972-431-2266 |
FX: 972-531-2266 | e-mail: bsherma1@...

-----Original Message-----
From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com]
Sent: Sunday, April 09, 2006 11:58 AM
To: TaxoCoP@yahoogroups.com
Subject: [TaxoCoP] Digest Number 241

There are 5 messages in this issue.

Topics in this digest:

       1. Re: Auto-categorization tools
            From: "Bhojaraju G" <bhojaraju.g@...>
       2. Auto-categorizing tools - a perfect topic for the wiki
            From: "Stephanie Lemieux"
<stephanieangelle.lemieux@...>
       3. RE: Auto-categorization tools
            From: "Foye, Chris" <chris.foye@...>
       4. RE: Auto-categorizing tools - a perfect topic for the wiki
            From: "Seth Earley" <seth@...>
       5. A member introduction, questions and a taxonomy resource
            From: "Alexandra" <amprose2001@...>


________________________________________________________________________
________________________________________________________________________

Message: 1
    Date: Sat, 08 Apr 2006 19:43:07 -0000
    From: "Bhojaraju G" <bhojaraju.g@...>
Subject: Re: Auto-categorization tools


Hi



I saw this tool, this is also good one.



Grokker has changed the search environment by using data clustering and
improved graphics.

http://www.grokker.com/ <http://www.grokker.com/>

I got results from Grokker Search like this:

https://beta.grokker.com/grokker.html?query=Bhojaraju&Yahoo=true&Wikiped
\
ia=true&numResults=250
<https://beta.grokker.com/grokker.html?query=Bhojaraju&Yahoo=true&Wikipe
\
dia=true&numResults=250>

Thanks and Regards

Bhojaraju G
Lead - Knowledge Management
************************************************************************
\
***********************************
KM Cyberary:   http://www.bhojarajug.freeservers.com/cyberary.html
<http://www.bhojarajug.freeservers.com/cyberary.html>
KM-Forum:       http://groups.yahoo.com/group/KM-Forum/
<http://groups.yahoo.com/group/KM-Forum/>
LinkedIn:         https://www.linkedin.com/in/bhojarajug
<https://www.linkedin.com/in/bhojarajug>
Skype Me:        bhojarajug

--- In TaxoCoP@yahoogroups.com, Linda Farmer <lfarmer@...> wrote:
>
> Hi Becky,
>
> Here is a list of auto-categorization tools which usually incorporate
> various degrees of taxonomy management. Some CMS vendors fold in
> auto-categorization functionality into their products (e.g. Stellent)
> but are not included here. This list includes only those tools which
> offer their auto-categorization/taxonomy management products as stand-

> alone or linkable to other applications through APIs. Schemalogic is
> not an auto-categorization tool but provides very powerful and
> comprehensive metadata and taxonomy management that serves up
> taxonomies and metadata to applications like auto-categorization and
> CMS.
>
> Could well be I've missed some auto-categorization/taxonomy tool out
> there. Those in the know, please let me know so we can build a
> complete list. Perhaps the list should be broadened to include other
> applications that incorporate significant auto-categorization and
> taxonomy management features into their product mix e.g. IBM, Inmagic
> products. Note: Stratify (http://www.stratify.com) is a major vendor
> of this type of tool but is now focusing on the legal domain.
>
>
>
> Auto-Categorization Tools
>
>
>
> Autonomy: http://www.autonomy.com (Includes Verity)
>
> Convera: http://www.convera.com
>
> Data Harmony: http://www.dataharmony.com
>
> Endeca: http://endeca.com
>
> Entrieva: http://www.entrieva.com
>
> Interwoven: http://www.interwoven.com
>
> Inxight: http://www.inxight.com
>
> Nstein Technologies: http://www.nstein.com
>
> SchemaLogic: http://schemalogic.com
>
> Siderean: http://www.siderean.com
>
> Smartlogik: http://www.aprsmartlogik.com
>
> Teragram: http://www.teragram.com
>
> Vivisimo: http://vivisimo.com
>
> Wordmap: http://www.wordmap.com
>
>
> Regards,
>
> Linda Farmer
>
> Second Knowledge Solutions
> http://k2s.ca
> email: lfarmer@...
> voice: 905-465-3387
>
>
> On 6-Apr-06, at 12:36 PM, Becky Sherman wrote:
> I wanted to know what products or tools do you currently have that
> help you manage your taxonomy? I heard someone talking about Data
> Harmony and wanted to know what other companies do this type of stuff
> - as I am trying to learn about what is out there because we don't
> have a really knowledge management area or tools in the enterprise -
> so data is everywhere. All help is good help - b




[This message contained attachments]



________________________________________________________________________
________________________________________________________________________

Message: 2
    Date: Sat, 08 Apr 2006 21:04:44 -0000
    From: "Stephanie Lemieux" <stephanieangelle.lemieux@...>
Subject: Auto-categorizing tools - a perfect topic for the wiki

Hi all,

Not to be a wiki-pusher, but I think this topic would lend itself
perfectly to elaboration on the wiki.
All of you who have mentioned tools or have specifications about what a
tool actually does, please visit the wiki and add a comment under the
tool (or add a new one to the list).
http://taxocop.wikispaces.com/TaxoTools

That way we can keep a running tab on all of this valuable content.

Thanks!
Stephanie






________________________________________________________________________
________________________________________________________________________

Message: 3
    Date: Sat, 8 Apr 2006 22:17:35 +0100
    From: "Foye, Chris" <chris.foye@...>
Subject: RE: Auto-categorization tools

All,

Apologies, Heather is right - I was responding to the point about tools
for taxonomy management as opposed to auto-categorization.

Brgds

Chris

________________________________

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf
Of Heather Hedden
Sent: 07 April 2006 18:25
To: TaxoCoP@yahoogroups.com
Subject: Re: [TaxoCoP] Auto-categorization tools


I use Synaptica quite a bit, and as far as I know there are no
auto-categorization functions.

This message thread had earlier included the topic of "tools that help
manage your taxonomy" in addition to more specific auto-categorization
tools. So, perhaps Chris was responding to the earlier query for "tools
that help you manage your taxonomy."

In either case, there is also a growing listing of tools on the new
Taxonomy Community of Practice Wiki:
http://taxocop.wikispaces.com/TaxoTools
Now we just need someone the categorize the list among those tools that
offer auto-categorization or other features.

--Heather

-----------------------------------------------------
Heather Hedden
Hedden Information Management
Carlisle, MA
Heather@...
http://www.Hedden-Information.com



[This message contained attachments]



________________________________________________________________________
________________________________________________________________________

Message: 4
    Date: Sat, 8 Apr 2006 19:41:31 -0400
    From: "Seth Earley" <seth@...>
Subject: RE: Auto-categorizing tools - a perfect topic for the wiki

You wiki pusher you...  ;-)

I agree.  And Stephanie you have been doing a great job on the wiki.  A
couple of people had offered to create synopses of CoP calls and post
them - I won't name names (yet...).

Perhaps I can also encourage others to stop in and post their profiles
on the wiki. If you have some thoughts on tools - feel free to add them.


(And Kate - now that I am spelling it correctly will you stop in?>
<smile>)

Here is the url http://taxocop.wikispaces.com/

Seth
-----------
Subject: [TaxoCoP] Auto-categorizing tools - a perfect topic for the
wiki

Hi all,

Not to be a wiki-pusher, but I think this topic would lend itself
perfectly to elaboration on the wiki.
All of you who have mentioned tools or have specifications about what a
tool actually does, please visit the wiki and add a comment under the
tool (or add a new one to the list).
http://taxocop.wikispaces.com/TaxoTools

That way we can keep a running tab on all of this valuable content.

Thanks!
Stephanie





________________________________________________________________________
________________________________________________________________________

Message: 5
    Date: Sun, 09 Apr 2006 15:47:52 -0000
    From: "Alexandra" <amprose2001@...>
Subject: A member introduction, questions and a taxonomy resource

Hello! I have been a lurker for the last year and was recently reminded
by Seth to introduce myself.

My name is Alexandra Proserpio from Sandia National Laboratories. I
support corporate classification efforts for search and browse (e.g.,
A-Z Index) applications. I also consult with SME and developers on IA
issues. My educational background is in library science.

A few years ago, we deployed a 2K node taxonomy (Verity's VIC) to
improve intranet search. It turned out that searchers only browsed the
taxonomy when the quality of search results met their needs which wasn't
often. So we refocused our efforts to improve the quality of our search
engine, developed best bets and a federated search and began qualitative
and quantitative testing. In a recent search satisfaction survey, 69% of
respondents indicated they find what they are looking for 75 to 100% of
the time. We are currently exploring the value of semantic technologies.
I hope to leverage the taxonomy (with many tweaks) if we decide to go
that route. On a related project, we are also exploring the value and
opportunities for enterprise taxonomy.

A few questions to members:

1) For those of you who lead/manage an enterprise taxonomy, are you
required to prove ROI? If so, how often and how do you do this?

2) In your organizations, do you have a (formal or informal) taxonomy
(or ontology) COP, COI or user groups?

Lastly, I'd like to share an out of print report I came across recently;
it may be of some interest for members. Check with an academic library
in your area.

Gilchrist, Alan and Peter Kibby. Taxonomies for business: access and
connectivity in a wired world. London. TFPL Ltd. 2000.  Here is a brief
snapshot of the report:
http://www.bokis.is/iod2001/papers/Gilchrist_paper.doc

Thanks. I look forward to contributing to the group!

Alexandra M. Proserpio






________________________________________________________________________
________________________________________________________________________



------------------------------------------------------------------------
Yahoo! Groups Links




------------------------------------------------------------------------

The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential and/or privileged
material.  If the reader of this message is not the intended recipient,
you are hereby notified that your access is unauthorized, and any review,
dissemination, distribution or copying of this message including any
attachments is strictly prohibited.   If you are not the intended
recipient, please contact the sender and delete the material from any
computer.

Thanks, Marcia.  Good pointer and interesting reading.  It also
contains some pointers to some additional interesting material.

I've intermittently looked at the searches that return zero results
and I'm not sure if they provide much insight to me, except for
people's inability to spell :-)

What I've found for these searches is: A) the percent of such searches
is a very stable 5-7% of searches during any given time period; B) the
number of distinct searches for any particular term that returns zero
results is usually less than 0.1% of searches; and C) that almost
every single such search contains a misspelling.  The search engine I
use doesn't do automatic correct of spelling, but it does suggest
alternative spellings based on a dictionary (the typical, "did you
mean xyz" kind of link).

It seems like the only real action that we might take from this would
be to monitor the "no result" searches for common misspellings and
maintain our synonym list with that insight.  It doesn't seem like
there is much we might do otherwise.

Is this unusual or am I not looking at this right?

Thanks again.

Lee


On 4/7/06, Marcia Morante <marcia@...> wrote:
>
>
> Lee -
>
> Check out Lou Rosenfeld's blog on the subject; a lot of  good discussion but
virtually nothing in the way of commercial products.-
http://www.louisrosenfeld.com/home/bloug_archive/000222.html
>
> One of the most revealing reports that you should be  able to get from your
search engine is "search terms with 0 results".  I  use this a lot to see the
language that people are using and what they  unsuccessfully tried to retrieve.
>
> Good luck,
>
> Marcia
>
> Marcia  Morante
> KCurve,  Inc.
> (718)881-5915 - office
> (917)821-2087 - mobile
> http://kcurve.com
>
>

Christine - Thanks a lot for the insight on the tool and how you've
used it (or not quite, anyway ;-) ).  I especially appreciate the
insight on the learning timeline - very insightful.

I'm looking for something a bit simpler, though, at the moment - but
I'll keep this in mind!

Thanks again!
Lee

On 4/10/06, Christine Connors <Christine_Connors@...> wrote:
> Lee -
>
> We have the same "very skinny" Long Tail on our Intranet.  There are
> products that "learn" from the user's queries and SERP selections-
> Recommendation Engines. Amazon has probably the most famous one of all -
> "people who bough xxx also bough yyy."
>
> We have one from Verity. (This Google search highlights some of those
> available:
>
<http://www.google.com/search?q=recommendation+engine&start=0&ie=utf-8&oe=utf-8&\
client=firefox-a&rls=org.mozilla:en-US:official>)
>
> Basically, it grabs data from the clickstream - the user's query, the
> results they opened, and the ones they "stayed with." If they use the
> browsable taxonomy as part of their search, it also learns which
> categories they used. It then can begin making recommendations as part of
> your UI. It can recommend any object-type that you have defined: document,
> category, person, etc.  It works, but the catch is you have to let the
> tool learn behind-the-scenes for a  while before the quality of
> recommendations is something you'd be willing to put into production. On
> large sites, that can take months. In fact, we didn't care for the results
> after 3 months, and then lost a team member and forgot about it. It's not
> in our production UI, but the recommendations on the Dev box are getting
> better every day!  :)  Another thing for the old To Do list...
>
> I feel a rant coming on about some other aspects of this thread... but
> I'll save that for another post!
>
> Cheers!
> Christine
>
>
>

I will second that request!  Please, please trim!

There have been at least two Digests recently that I finally just gave up on
trying to read, because there were two or three COMPLETE copies of previous
Digests included, and I couldn't figure out which was the new message and
which was the old without reading through each to see if I'd already read
it.  And frankly, I simply don't have the *time* to do that!

Thank you, Seth, for suggesting this!  and a BIG thank you to all who
remember to do it!

Grace Judson
www.svahaconcepts.com

---

From: TaxoCoP@yahoogroups.com [mailto:TaxoCoP@yahoogroups.com] On Behalf Of Christine Connors
Sent: Monday, April 10, 2006 8:48 AM
To: TaxoCoP@yahoogroups.com
Subject: Re: [TaxoCoP] Auto-categorization of search terms?

Lee -

We have the same "very skinny" Long Tail on our Intranet.  There are
products that "learn" from the user's queries and SERP selections-
Recommendation Engines. Amazon has probably the most famous one of all -
"people who bough xxx also bough yyy."

We have one from Verity. (This Google search highlights some of those
available:
<http://www.google.com/search?q=recommendation+engine&start=0&ie=utf-8&oe=utf-8&client=firefox-a&rls=org.mozilla:en-US:official>)

Basically, it grabs data from the clickstream - the user's query, the
results they opened, and the ones they "stayed with." If they use the
browsable taxonomy as part of their search, it also learns which
categories they used. It then can begin making recommendations as part of
your UI. It can recommend any object-type that you have defined: document,
category, person, etc.  It works, but the catch is you have to let the
tool learn behind-the-scenes for a  while before the quality of
recommendations is something you'd be willing to put into production. On
large sites, that can take months. In fact, we didn't care for the results
after 3 months, and then lost a team member and forgot about it. It's not
in our production UI, but the recommendations on the Dev box are getting
better every day!  :)  Another thing for the old To Do list...

I feel a rant coming on about some other aspects of this thread... but
I'll save that for another post!

Cheers!
Christine



>On our intranet, we use a search engine that produces weekly reports
>of the terms that are used.  Like many search engines, the reporting
>is based on the exact terms used by the users - right down to the case
>of the terms (though the engine itself is case insensitive).

>What I've found is that even the most common searches are a
v>anishingly small percentage of the overall set of search terms - the
>most common search term is commonly around 1% of search terms and the
>top 100 terms usually are about 20% of all searches within a week.

>I've taken a stab at reviewing the search terms and doing some manual
>categorization to start to "see the forest for the trees", but it's
>tedious and time-consuming and, especially when you get another bunch
>every week, hard to keep up with.







Christine JM Connors
Metadata Architect

Raytheon Company
Information Technology
600 Technology Park Drive
M/S RTN-2S119
P.O. Box 7038
Billerica, MA 01821
978.436.3195 voice
978.436.3205 fax
508.314.2824 cell Christine_Connors@...

Hi Marcia -

No, I haven't tried Autonomy's yet. I suppose I will sooner or later. Is
anyone using Autonomy's? If so - please share!

Of course, this still doesn't solve the zero-hits (or one-two hits) per
query problem... back to that Long Tail.

Christine





Christine JM Connors
Metadata Architect

Raytheon Company
Information Technology
600 Technology Park Drive
M/S RTN-2S119
P.O. Box 7038
Billerica, MA 01821
978.436.3195 voice
978.436.3205 fax
508.314.2824 cell Christine_Connors@...



TaxoCoP@yahoogroups.com wrote on 04/10/2006 02:00:23 PM:

> Hi Christine -
>
> Just curious about whether you tried the Autonomy similarity feature
> and how the one from Verity compares.  Does it use the system ID or
> do you or will you require a separate login?
>
> Thanks,
>
> mm
>
> Marcia Morante
> KCurve, Inc.
> (718)881-5915 - office
> (917)821-2087 - mobile
> http://kcurve.com
>
>

In message <20060408005530.D482D405568@...> on Fri,
7 Apr 2006, Seth Earley <seth@...> wrote
>I just received a note from a CoP member suggesting that I remind
>people to trim posts - especially when responding to a digest.

Can I say amen to that, and add the plea that people should ensure that
the subject lines of their posts give a clear indication of the content
- surely a superfluous request in a taxonomy group, but I have noticed
some which retain an uninformative heading such as "Digest Number 241".
:-)

It is helpful to be able to scan down a list of messages and know which
are likely to be of enough interest to open and read.

Also, please use the "reply" or "follow-up" function when continuing a
discussion thread, and the "start new thread" function when starting a
new topic. That will help those of us whose email software builds
threads based on "References:" or "In-Reply-To:" headers irrespective of
changes in the subject line.

Leonard Will
--
Willpower Information       (Partners: Dr Leonard D Will, Sheena E Will)
Information Management Consultants              Tel: +44 (0)20 8372 0092
27 Calshot Way, Enfield, Middlesex EN2 7BQ, UK. Fax: +44 (0)870 051 7276
L.Will@...               Sheena.Will@...
---------------- <URL:http://www.willpowerinfo.co.uk/> -----------------