Oracle Text Retrieval Product (TextServer3)
Date: 1995/04/07
Message-ID: <3m32e4$54c_at_dcsun4.us.oracle.com>
I noticed a couple of recent mails about availability of text search products for Oracle, and thought I would pass on the following info for anyone interested.....
Received:
From: An Information Service of INDIVIDUAL Inc. <UKOO1:OASUN1.US.ORACLE.COM: To: firstois_at_us.oracle.com
Subject: INFORMATION HIGHWAY OR INFORMATION OCEAN? ORACLE DELIVERS (...) Profid: ORACLE11
Addrid: DEPT1
SUBJECT: INFORMATION HIGHWAY OR INFORMATION OCEAN? ORACLE DELIVERS SOFTWARE TO CONTROL THE FLOOD OF TEXT-BASED INFORMATION SOURCE: PR Newswire via First! by Individual, Inc. DATE: April 3, 1995
INDEX: [7]
FLORENCE, Italy, April 3 /PRNewswire/ via First! -- The promise of the information highway is a continuous flow of information -- text, audio and video -- to the desktop. The reality is a sea of information, with islands of non-integrated text, images and relational data coexisting, but not working together. Oracle TextServer tackles information overload by integrating textual and relational data to deliver the most useful and relevant information to the user.
At its European Oracle User Group (EOUG) conference, Oracle Corp. (NasdaqNNM: ORCL) today announced the immediate availability of Oracle TextServer3 with ConText, the first software product to intelligently search and manage text as easily as relational data. The software stores, analyzes and retrieves electronic documents by their content or actual meaning.
ConText, Oracle's revolutionary natural language processing technology, uses its intelligence to understand the meaning of documents, weeding through scores of irrelevant text to quickly deliver precise results. ConText's unique capabilities enable searching by "theme" as well as "speedreading" and automatic summarization.
"Today's business environment requires quick response, and the sheer
volume of information is overwhelming, rendering much of it useless because
it's difficult to process," said Farzad Dibachi, vice president of Oracle's
New Media Division. "Oracle TextServer is another building block in
Oracle's plan to deliver an integrated infrastructure for the fast, easy
management of all data types."
Scalable Document Repository
Industry experts agree that the ConText/RDBMS integration is a landmark breakthrough in integrated information management. Based on a scalable client/server architecture, Oracle TextServer3 with ConText takes full advantage of Oracle7's parallel server technology to deliver the most efficient text retrieval system available. Built to support parallel index
and query operations, it can easily handle multi- gigabyte, distributed text databases with millions of documents.
Smart Searches
Oracle TextServer3 uses a dictionary of more than 600,000 words and extends text retrieval beyond the bounds of current solutions. In addition to Boolean and probabilistic word and phrase searches, Oracle TextServer3 employs ConText to provide thematic content-oriented search capabilities. Thematic search analyzes the document content instead of counting word occurrences. Search terms need not even appear in the document.
For example, in a traditional text-retrieval system, a search for "terminals," would retrieve all documents containing the word "terminal," regardless of the context in which the word is used. More intelligently, since Oracle ConText understands word connotations, it distinguishes when "terminal" is used in the sense of computer terminals, train or bus terminals, or terminal illness and retrieves the most relevant documents accordingly. Equally important is ConText's ability to detect themes in a document. Hence, documents discussing VT220s would be retrieved by a query for "terminal," even if the word "terminal" is not mentioned.
Automatic Summarizing and Indexing
In Oracle TextServer3, ConText automatically creates efficient key- word indices and provides a range of document summary formats. Index volume can be reduced by indexing only those words bearing thematic importance. Query performance is improved and only the most relevant documents are retrieved. Retrieved documents can be viewed in their original WYSIWYG form, edited in their corresponding word processor or highlighted by ConText for speedreading.
For the first time, users can drill down to the meaning within documents. Using the SpeedRead viewer, a document can be automatically summarized to various levels of reduction: Full (100 percent), Read (50-60 percent) and
Glance (20-30 percent). Alternatively, SpeedRead can highlight only those
sections of the document that contribute to understanding its meaning. In either viewing mode, users can browse through large document repositories and quickly identify the sections of interest. Productivity is increased and knowledge is more easily gained.
"With the ever-expanding amount of electronic information flowing through
computer networks, we need to restore the balance between senders and
receivers," said Geoffrey Bock of the Patricia Seybold Group. "With
technologies such as ConText, users can focus on business processes and
business activities rather than on systems processes and computational
activities."
"Oracle TextServer3 with ConText will replace two separate management and
retrieval systems for text and structured data," said Warwick Cathro,
assistant director-general of the National Library of Australia. "Now our
users in Australia and New Zealand will be able to efficiently and
transparently access data from any information source -- be it book,
newspaper, magazine, academic paper or any other printed document."
/CONTACT: Eve Kowtko Smith or Dawn Echols of Oracle, 415-506-4176/ (ORCL)
[04-03-95 at 08:10 EDT, PR Newswire, File: p0403081.001] Press Return to read next message or Cancel to close.
Press Return to read next message or Cancel to close. gingram.UK <Detail> <Insert>
-- Geoff Ingram +-----------------------------------------------+ | Internet email address: gingram_at_uk.oracle.com | +-----------------------------------------------+Received on Fri Apr 07 1995 - 00:00:00 CEST