Hello Yuri.
Понедельник Сентябрь 15 1997 18:00, Yuri Kuleshov wrote to All:
Вот месяца два назад натаскал по data mining.
=== Cut ===
Из URL: ftp://ftp.sas.com/pub/neural/FAQ.html
Maintainer: saswss@unx.sas.com (Warren S. Sarle)
----------------------------------------------------------------
Subject: [32] Data Mining
Introductions:
W. Frawley, G. Piatetsky-Shapiro, and C. Matheus, "Knowledge Discovery in
Databases: An Overview", AI Magazine, pages 213-228, Fall 1992.
Article Collections:
International Journal of Intelligent Systems 7(7), September 1992.
IEEE Transactions on Knowledge and Data Engineering 5(6), December 1993.
Gregory Piatetsky-Shapiro and William J. Frawley, editors, "Knowledge
Discovery in Databases", MIT Press, 1991.
----------------------------------------------------------------
Subject: [5-1] FTP Repositories
CMU AI Repository:
The CMU Artificial Intelligence Repository was established by
Carnegie Mellon University to contain public domain and freely
distributable software, publications, and other materials of
interest to AI researchers, educators, students, and practitioners.
The AI Repository currently contains more than a gigabyte of
material and is growing steadily.
The AI Repository is accessible by anonymous FTP from
ftp.cs.cmu.edu:/user/ai/ [128.2.206.173]
by AFS (Andrew File System) from
/afs/cs.cmu.edu/project/ai-repository/ai/
and by WWW from the URL
http://www.cs.cmu.edu/Web/Groups/AI/html/repository.html
Be sure to read the files 0.doc and readme.txt in this directory.
discover/ Discovery and Data-Mining
Data Mining and Knowledge Discovery in Databases:
Explora is a data mining package written in Lisp for the Macintosh. It
includes a natural language hypertext-type interface for presentation
of dicoveries. It is available by anonymous FTP from
ftp.gmd.de:/GMD/explora/
as the files Explora.sit.hqx and README. For more information, see
http://orgwis.gmd.de:80/explora/
INSPECT is a PC-based data mining tool with visualization and neural
networks. It is available by anonymous FTP from
ftp.tuwien.ac.at:/Sources/NeuralNet/Inst-of-Chem/
See the readme.txt file for details.
Data Mining and Knowledge Discovery in Databases:
Data Mine
http://www.cs.bham.ac.uk/~anp/TheDataMine.html
[Bibliographies, On-line papers, Software, and Other Resources]
Andy Pryke <anp@cs.bham.ac.uk>
Knowledge Discovery Mine
http://info.gte.com/~kdd/
ftp.gte.com:/pub/kdd/
[The Knowledge Discovery Mine includes the KDD FAQ, a catalog of
commercial and public domain tools for discovery in data, workshop
reports, as well as back issues of the KDD-Nugget mailing list. It
also includes a list of AI resources compiled by Chris Matheus
<cjm1%scintilla@gte.com>]
Gregory Piatetsky-Shapiro <gps@gte.com>
Data Mining:
Also known as Knowledge Discovery in Databases (KDD) was been defined
as "The nontrivial extraction of implicit, previously unknown, and
potentially useful information from data" in Frawley and
Piatetsky-Shapiro's overview. It uses machine learning, statistical
and visualization techniques to discover and present knowledge in a
form which is easily comprehensible to humans.
----------------------------------------------------------------
Subject: [1-6] Commercial Expert System Shells
Angoss Knowledge Seeker. Angoss is a data-mining tool that can be used
to produce knowledge bases of rules by infering cause and effect
relationships from a database. The DOS version costs $799 and the
Windows version $899. For more information, write to Angoss Software
International Ltd., 430 King Street W., Suite 201, Toronto M5V 1J5,
Canada, call 416-593-1122, or fax 416-593-5077.
PowerSMARTS (multimedia rule-based logic), DataSMARTS (database mining),
NeuroSMARTS (neural net) for Macintosh or PC Windows. Costs range from
$495 to $4995. Cognition Technology Corporation, 1000 Massachusetts Ave,
Cambridge, MA 02138, 617-492-0246, FAX 617-492-0247, Internet ctcma@aol.com
А вот дока по пакету:
The NETMAP Software System
[Image]
Reviewed by Dave Baize, special agent supervisor, Arizona Office of Attorney
General. The views expressed are those of the author and do not necessarily
reflect those of the State of Arizona or the Arizona Attorney General.
[Image] MANUFACTURER: ALTA Analytics Inc.
555 Metro Place North, Suite 175
Dublin, OH 43017
(614) 792-2222, (614) 792-3694 fax
Email: infomaster@alta-oh.com
http://www.alta-oh.com
[Image]PLATFORMS: NETMAP runs on HPUX 9.0 or greater, EBMAIX 3.2, DEC Alpha
(OSF/1) 3.2 or greater, DEC ULTRIX (RISC) 4.2 or greater, Sun Solaris 2.4 or
Sun OS 4.1.x using XWindow System Version 11, Release 4 or later, and
OSF/Motif Release 1. 1 or later. Support for Windows NT will be forthcoming.
NETMAP supports HPGL/2 and PostScript output devices.
[Image]BACKGROUND: The objective of the Office of Attorney General was to
find and prosecute racketeering through evidence of money laundering
schemes. In 1988 we launched an effort to obtain data from the U.S.
Department of the Treasury. The Office requested all Currency Transaction
Reports and Currency and Monetary Instrument Reports containing an Arizona
zip code, or filed at an Arizona institution or port of entry. We now
receive data from a variety of sources--the U.S. Treasury Department, local
institutions and businesses, phone data, motor vehicle data, search warrant
proceeds, assets (real property, boats, planes, etc.) and internal data.
[Image]PRODUCT FUNCTIONALITY: NETMAP allows advanced visual data analysis
and data mining of multiple and disparate data sources. The first "real"
test of NETMAP was a large financial case involving tax fraud. The
investigating agency believed one large enterprise was involved and the
investigation was structured with that theory in mind. NETMAP, however,
vividly showed that there was not one large enterprise but at least three
smaller ones. These enterprises may have had other relationships (social,
family, etc.) but, financially, they were not connected. There were three
separate leaders, two of whom had been charged and a third who had not. The
benefits of such insight to investigative and prosecutorial planning is
obvious.
[Image]STRENGTHS: NETMAP allows me to perform complex queries of data from
multiple disparate sources. A common algorithm to target money laundering
schemes might be: "give me all transactions from Santa Cruz County for the
year 1992 where the amount was less than $10,000 and there were multiple
transactions within a week the sum of which were over $10,000."
I have been very pleased with NETMAP's performance. The capabilities of the
NETMAP System never cease to amaze me. It is extremely versatile and data
insensitive. The system is limited only by the user's imagination.
[Image]WEAKNESSES: I found some initial weaknesses when NETMAP was first
installed in 1990, however, most of them have been addressed by later
releases of the software. There is still a limitation in the amount of data
that can be analyzed at any one time. We would like to see a terabyte of
data be able to be used.
[Image]PRODUCT CHARACTERISTICS: Visual representation is the real power of
NETMAP. Data can be quickly and intuitively viewed in a variety of
chart/graph formats providing different perspectives for analysis. NETMAP is
accurate, reliable and easy to use. We are able to do more on the cases on
which we work. Where in the past we did not have the resources to fully
analyze and chart out a specific case, we often do now. NETMAP provides
additional leads, shows who may or may not be important, and can offer
investigative direction through its various "what if" functions. ALTA
Analytics Inc. offers four basic NETMAP courses: User Training, NETMAP
Presentation Tool (NPT), System Administration and Application Development.
Six days are required to complete all of the above courses. The only
prerequisite for NETMAP Training is a UNIX System Basics course. Online help
is provided.
[Image]PRODUCT DELIVERABLES: NETMAP's utilities allow data to be subjected
to advanced grouping and analysis routines. Common connections lets analysts
discover critical associations by charting all relationships to an entity;
emergent groups reveal hidden organizational structures; and step links
displays the shortest path of connections between common groups. We have
often used NETMAP graphics as evidence in court.
[Image]SELECTION CRITERIA: Our original objective was to obtain a commercial
PC database manager to manipulate complex sets of data. While the original
software did an excellent job of indexing the data and allowing relatively
quick searches for simple data structures, such as a person's name, a
business, an address or an account number, the system grounded to a halt
when faced with more complex queries.
In the fall of 1991, the Rocky Mountain Information Network (RMIN) let me
observe a graphical analysis system called TTollmap, loaded with telephone
toll records. TTollmap identified the various relationships among the
numbers, times, call duration, etc. and graphically depicted the structure
of the organization represented. One could instantly see which numbers were
important and how the members of the organization interrelated. Even more
important, TTollmap allowed "whatif" questions to be posed. RMIN referred me
to the developer of the TTollmap system--ALTA Analytics in Dublin, Ohio. I
learned that the TTollmap program had beengreatly expanded beyond the RMIN
system and was now called NETMAP.
[Image]VENDOR SUPPORT: ALTA Analytics installed NETMAP at our office along
with a custom-designed front end. This enabled us to process virtually any
data set that shows relationships among the individual members. Through the
purchase of a yearly maintenance program, ALTA provides software updates and
telephone support.
=== Cut ===
Могу продолжить еще на 20 кил zip, если очень надо:)
Andrew
-- end of forwarded message --
--
Kittle