Message-ID: <1999322.1075848226462.JavaMail.evans@thyme> Date: Tue, 16 Jan 2001 09:32:00 -0800 (PST) From: marie.hejka@enron.com To: steven.kean@enron.com Subject: Your thoughts required by 3 :00 p.m. 1/17 meeting. Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-From: Marie Hejka X-To: Steven J Kean X-cc: X-bcc: X-Folder: \Steven_Kean_June2001_5\Notes Folders\Knowledge management X-Origin: KEAN-S X-FileName: skean.nsf ----- Forwarded by Marie Hejka/Corp/Enron on 01/16/2001 05:32 PM ----- Marie Hejka 01/16/2001 12:05 PM To: Melissa Becker/Corp/Enron@ENRON, Anthony Mends/Enron Communications, Andrea Yowman/Corp/Enron@ENRON, Joe Wong/enron@enronXgate@enronXgate, Debbie R Brackett/HOU/ECT@ECT, Pegi Newhouse/HOU/EES@EES, John Gillespie/Enron@EnronXGate, Kathleen Pope-Sance/HOU/EES@EES, Steve Woods/EPSC/HOU/ECT@ECT, Georgeanna Hoiseth/Corp/Enron@ENRON, Judith Schlesinger/HOU/ECT@ECT, Allan Sommer/Corp/Enron@Enron cc: Beth Perlman/enron@EnronXgate, George Wasaff/Enron@EnronXGate, Kent Morrison/NA/Enron@ENRON, John Simmons/NA/Enron@Enron, Paul Timberlake/ET&S/Enron@Enron, Allen Elliott/HOU/ECT@ECT Subject: Your thoughts required by 3 :00 p.m. 1/17 meeting. Please read and be prepared to discuss at tomorrow's task force meeting. (Paul is scheduled to speak to us to help us make a decision. FYI - I plan to circulate the agenda before the meeting.) Below are Paul Timberlake's notes from Friday morning's Software Selection Team meeting. You may remember in the beginning of this project we discussed piloting Autonomy or an unstructured data management tool in order to reach qualitative and quantitative information. Some discussion was lent to finding a product perhaps which would search email to identify experts. Further, we discussed the value in finding a product which could help us identify experts within Enron. Since these discussions, we hired a Business Analyst who interviewed some 29 people and surveyed 30 additional people who concurred finding experts in Enron would be useful. However, only 50% of those interviewed suggested they would tag email for others to search regardless if it could help identify experts. But, over 90% of those surveyed agreed an Enterprise Search Engine would be useful. Based on what we now know, the Software Selection Team will be evaluating three Enterprise Search Engine vendors (Inktomi, Autonomy, and Verity). These vendors DO NOT search email like the vendor TACIT. Autonomy provides some feature/functionality to search email (the full extent of which will be evaluated by the SST). Are we prepared to invest in a much more expensive search engine with categorization feature/functionality which may provide a more robust search platform of the future or should we decide to invest in the best value for our current needs? Categorization is the classification of information sources, such as documents or Web pages into a taxonomy. With some products, the taxonomy must be determined beforehand and rules specified (both manual operations) to tell the product how to classify information sources it encounters. Other tools claim they create the taxonomy using a proprietary method. See Paul's "Issue" section below. ----- Forwarded by Marie Hejka/Corp/Enron on 01/12/2001 05:52 PM ----- Paul Timberlake 01/12/2001 04:16 PM To: Marie Hejka/Corp/Enron@ENRON, Kent Morrison/NA/Enron@ENRON, Allen Elliott/HOU/ECT@ECT cc: John Simmons/NA/Enron@Enron Subject: KM/IM Meeting Notes 1/12/2001 Meeting Notes From the 1/05 weekly meeting, it was decided that the tool being sought was best classified as an Enterprise Search Engine (ESE) solution. This distinction was arrived at after reviewing the characteristics described in the charter and associated survey as to what was meant by an unstructured data management tool. Three leading vendors in this market are Inktomi, Autonomy, and Verity. To stay within the time frame of the overall project, the scope of the search tool evaluation will focus on these three vendors. Evaluation criteria have been developed to compare the tools. This criteria focuses on the tools' indexing and search functionality At present the criteria excludes categorization functionality. The reason for this is categorization is expected to involve manual effort from various groups who own content that will be categorized. This characteristic is at odds with the project charter requirement that the tool be unobtrusive to current processes and culture. Because indexing and search functionality is becoming similar among ESE tools, it is expected that the ESE evaluation will boil down to a price decision with Inktomi being the winner. The three ESE vendors are being lined up to visit Enron and present their products beginning next week through the week of January 22nd. All interested parties are encouraged to attend. Issue At issue is whether categorization functionality should be included in the ESE evaluation - or continue to focus only on indexing and search functionality. All three vendors being evaluated include a categorization component. By not including categorization, we run the risk of selecting a vendor whose categorization capabilities may not be as robust as another's. Integrating another vendor's solution later on may prove difficult and costly. By including categorization functionality, the evaluation criteria become ambiguous. This is because the requirements around categorization are not well defined at present. Consequently, we run the risk of selecting an over-enginneered and most costly product whose perceived benefits may never be realized. We need more direction from the project sponsors as to what priority to place on categorization functionality. If it is considered a priority then we also need end user groups with specific problems identified that, if solved, could effectively evaluate the tool and justify any additional costs.