NetOwl TM Server 
IsoQuest, Inc. 
3900 Jermantown Road 
Suite 400 
Fairfax, VA 22030 
1. Introduction 
NetOwl TM Server is a powerful text analysis 
software product developed by IsoQuest, Inc. It 
supports business intelligence by providing fast and 
easy access to information stored on local Intranets 
and the global Internet. It organizes, analyzes, and 
summarizes data extracted by NameTag TM and any 
Full-Text search engine. It then presents the data for 
either searching or browsing. NameTag is a data 
extraction and indexing tool that finds proper names 
and other defined entities within an input text stream. 
NetOwl is the total application built on the NameTag 
core engine. NetOwl tags each desired document or 
Web page by person, organization, location, 
relationship and description giving you a browsable 
"back-of-the-book index.'" This allows for targeting 
exactly key information that is sought. With NetOwl 
it is no longer necessary to scroll through massive 
amounts of text to find exacdy what is useful. 
2. Search Functionality 
NetOwl is a server-based program which generates 
standard CGI commands that can be executed by any 
standard Web browser. NetOwl's Loader ingests 
documents from any text or HTML-based data source, 
sequentially loads the documents, then passes them 
on to NameTag and the full-text search engine for data 
extraction. The full-text search engine data is stored in 
a proprietary database while NameTag's exlracted data 
is stored in any ODBC-compliant database. 
There are two major functional areas of the NetOwl 
system. The NetOwl Server system consists off 
• Loader 
• Full-Text Search 
• NameTag 
• Database interface 
• Client 
The second NetOwl functional area comprises 
customization, maintenance and monitoring tools 
which provide support to the underlying processes of 
NetOwl: 
• NetOwl Administrator Tool 
• NetOwl Service Manager 
Both major areas use the relational database for 
information storage and retrieval. This is the only 
link to each of the functional areas (i.e., all common 
information is accessed from the central database). 
The major functional areas never talk direcdy to each 
other or require or pass information between each 
other. 
Loader 
The Loader component of the NetOwl server collects 
information to process (by crawling specified sites, 
USENET news groups and/or locally accessible files) 
and then creates an index of that data in the NetOwl 
database. 
Full-Text Search 
This Full-Text Search (P-TS) component accepts text 
documents from the loader and creates an index of all 
relevant words. The index thus created can later be 
searched by the client component to provide a list of 
relevant documents given a set of keywords. 
Database 
The Database component stores the index information 
generated by the Loader/NameTag combination. The 
database is a relational database linking instances of 
names to relationships, descriptions, aliases, ",rod 
documents. NetOwl does not supply the database. It 
does, however, supply the interface to a database 
through an ODBC layer so that any RDBMS system 
can be used (e.g., Access, SQL Server, Oracle, 
FoxPro, Informix, Sybase). 
Client 
The term "client" refers to the user interface you will 
use to browse NetOwl's visual index, display 
summary information, and access information that is 
most salient. The client component is either a CGI- 
15 
compliant program or an HTFP server extension 
which accepts requests transmitted from your WWW 
browser and returns an HTML page with the results 
of the request. The client is the vehicle used to access 
index information in the database. 
The GUI interface provides you with direct access into 
the NetOwl screens where you can design, submit, 
save, load and delete queries. It then communicates 
these queries to the database components and presents 
you with the full document text along with a glossary 
of the names, descriptions, relationships and words 
found by NameTag and the fiTS engine. 
NameTag 
NameTag represents the key technology behind 
NetOwl's indexing ability. NameTag is a data 
extraction engine that identifies and interprets key 
elements of free text, particularly names of people, 
organizations and locations. For each document it 
reads, NameTag builds a signature that includes each 
name or concept, where it's found in the document, 
and how it's interpreted. For example, "Wa,~ington" 
can be a city, state, last name, or part of a company 
name, depending on the context. "Sun" can be a last 
name, one of several companies, or a celestial object. 
NameTag identifies these pieces of information, helps 
to interpret them and eliminate the ambiguity, 
pulls in associated information such as ticker 
symbols for companies and variations of person 
names. The technology behind NameTag itself is a 
combination of a powerful pattern-matching technique 
known as finite state interpretation with a carefully- 
designed knowledge base. 
3. Further Information 
For more information contact: 
IsoQuest, Inc. 
3900 Jermantown Road 
Suite 400 
Fairfax, VA 22030 
Tel.: (703) 293-2350 
Fax: (703) 293-2353 
e-mail: isoquest@isoquest.com 
www.isoquest.com 
This technical description is condensed from The 
Delnhi Intranet Report, © 1996, The Delphi 
Consulting Group 
16 
