Enron email

Algorithmic Indexing

5 Jan 2001, 16:15 UTC

Software Selection Team Meeting Notes 1/05/01

From
Marie Hejka marie.hejka@enron.com
To
Melissa Becker, Andrea Yowman, Anthony Mends, Kathleen Pope-Sance, Joe Wong, Debbie R Brackett, pegi.newhouse@enron.com, John Gillespie, steve.woods@enron.com, Beth Perlman
Cc
George Wasaff, Steven J Kean, john.simmons@enron.com, paul.timberlake@enron.com
Bcc
George Wasaff, Steven J Kean, john.simmons@enron.com, paul.timberlake@enron.com
Date
Fri, 5 Jan 2001 08:15:00 -0800 (PST)
Below are this morning's meeting notes defining the scope of the Software 
Selection Team.  Remember the SST will determine what UDM/Enterprise Search 
Engine the KM/IM task force will pilot.

Meeting Attendees:
Marie Hejka
Georgianna Hoiseth
John Simmons
Paul Timberlake
Steve Woods
Quoted or forwarded from earlier messages
----- Forwarded by Marie Hejka/Corp/Enron on 01/05/2001 04:06 PM -----

	John Simmons
	01/05/2001 11:56 AM
		 
		 To: Marie Hejka/Corp/Enron@ENRON
		 cc: Paul Timberlake/ET&S/Enron@Enron
		 Subject: Update on KM - Enterprise Search Engine Approach

New Approach:

Instead of focusing on UDM (Unstructured Data Management), we focus on an 
Enterprise Search Engine, of which handling unstructured data is a feature 
that we may or may not include as a "must have" in the requirements. 
Create a document that lists the "must-have" features vs. the "nice-to-have" 
features, focusing on the capabilities of search engines. This list of 
features, when compared with the capabilities of a small set of 
enterprise-level search engines, will drive the ultimate selection.

Things to consider:

TIBCO provides Enron with an enterprise license for Portal Builder, our tool 
of choice for portals. TIBCO is leaning toward building interfaces with 
Verity, a popular search engine.
One characteristic of search engines, categorization, should not be 
considered a "must-have" requirement. The supposition is that business 
groups, like Legal or Contract Administration, which produce structured 
content, would be well-served by LiveLink. Rather, the Enterprise Search 
Engine (ESE?) would be driven more by the scenario where an Enronite would 
jump to his search engine, enter a keyword or words, and sift through the 
results, without having to know a lot about categories.
Enron is presently using Ultraseek and Altavista as search engines with a 
limited license.
As for unstructured data, such as files in a common drive, it will take some 
level of effort and design to make such data available to an ESE. We should 
therefore anticipate that an IT/Business group combination will have to own 
the process of working with individual departments to "search-enable" their 
data. 
A scenario for a pilot would be to use an existing portal and embed one or 
more ESE's and have it address LiveLink data, Intranet websites, and a couple 
of departmental data stores (such as Facilities), AND the Internet.
I propose that the scope of engines include Autonomy, Verity and Ultraseek. 
This may change, depending new information.