Verity Search Engine

Subject: Verity Search Engine
From: "Fish, Paula" <paula -dot- fish -at- cplc -dot- com>
To: 'TECHWR-L' <techwr-l -at- lists -dot- raycomm -dot- com>
Date: Mon, 26 Jun 2000 10:58:30 -0400

Does anyone have experience using the Verity search engine with an Adobe PDF
collection of documents? I have searched both the Verity and the Adobe web
sites looking more for tips on how the two products work in tandem. I have
not yet found what I am looking for as far as tips to enhance search

I work in a Document Center that has a collection of approximately 2,000
policies, procedures and related forms that are published on the corporate
intranet for employees in hundreds of location across two states. By the
end of this year, the audience of users will encompass a newly organized
Holding Company structure with entities in Florida as well as the flagship
business in North Carolina.

Since we started the electronic document center four years ago, it has been
a constant challenge to find strategies that support and provide a
successful method of document retrieval for the skill sets of our varied
customer base. Skill sets range from expertise skills in on-line
documentation and searching employees who do not know how
to navigate the corporate intranet much less construct queries at the
Corporate Document Program search interface that can help them find what
they are searching for.

We currently provide pre-coordinated indexing of the collection by function
(i.e. safety, environmental); collection (Employee Handbook; Benefit
Booklets); and grouping/impacted audience (general corporate population;
line repair procedures for the Transmission Department). What we have found
is that customers tend to not know and/or recognize the information by
indexing terms used.

We also offer full text searching capability. What we have found is that
not many customers know how to construct even the simplest queries using
Boolean logic. Random full text searching tends to bring numerous and
copious results. "Car Accident" for instance will pick up every word that
has "dent" in the word. Stemming, as explained, causes this but can not be
turned off. Rank and relevance is offered with the results listing but the
results are questionable. Upper case/lower (mixed) case cause negative
results. The exact title of a document can be keyed and it will appear as
no. 7 on a list of 20 hits; this same document may be ranked at a percent
much less than 100%. Are there features that can be tweaked in the
programming or interface between Verity and Adobe to minimize such results.
Or, another way to state the maximize the results?

Usability survey:
The Policies and Procedures team is in the process of developing a customer
usability survey.....can we assess the search method(s) our customers are
using to locate information? What limitations can our program address;
what is beyond our control? Does anyone have experience or recommendation
in how to proceed with this effort? Are there lessons learned about
constructing a usability survey? About administering the survey instrument?
About analyzing such resulting data? Implementation of any resulting action

Paula H. Fish
Corporate Document Center
(919) 546-4708

Previous by Author: chart navigators
Next by Author: RE: Should we hire this guy?
Previous by Thread: Re: Frame and macros?
Next by Thread: Re:FUN: question about recruiters (plus a little venting)

What this post helpful? Share it with friends and colleagues:

Sponsored Ads