--
Discovery community, focusing on the latest research and applications.
Contributions are most welcome and should be emailed,
with a DESCRIPTIVE subject line (and a URL, when available) to (kdd@gte.com).
E-mail add/delete requests to (kdd-request@gte.com).
Nuggets frequency is approximately weekly.
Back issues of Nuggets, a catalog of S*i*ftware (data mining tools),
and a wealth of other information on Data Mining and Knowledge Discovery
is available at Knowledge Discovery Mine site, URL http://info.gte.com/~kdd.
-- Gregory Piatetsky-Shapiro (moderator)
********************* Official disclaimer ***********************************
* All opinions expressed herein are those of the writers (or the moderator) *
* and not necessarily of their respective employers (or GTE Laboratories) *
*****************************************************************************
~~~~~~~~~~~~ Quotable Quote ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
If the fool would persist in his folly he would become wise.
William Blake (1757-1827),
thanks to http://www.mcs.net/~jorn/html/blake.html
Previous1NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Thu, 8 Aug 96 11:32:54 CDT
To: kdd@gte.com
From: 'Gerry McKiernan' (JL.GJM@ISUMVS.IASTATE.EDU)
Subject: Project Aristotle(sm)
Content-Length: 1600
Project Aristotle(sm)
Automated Categorization of Web Resources
I am pleased to announce the establishment of Project Aristotle(sm),
a clearinghouse or projects and research devoted to the automated
categorization of Web resources. The URL for Project Aristotle(sm) is:
For each project, it's name, if known, principal investigator,
project description, and relevant citations are provided. A hotlink
to an available demonstration or prototype is also provided, if
available. Entries are organized alphabetically by the name of the
organization with which the principal investigator is affiliated.
I am greatly interesting in developing this clearinghouse further
and would very much appreciate the name, e-mail and/or URL of similar
projects or investigations.
Presently, I am only interested in projects and prototypes that have
_applied_ filtering systems, text extraction and/or categorization, or
agents, robots or machine learning to the categorization of Web resources.
I am _not_ presently interested in work that reviews these approaches or
technologies in general.
I am particularly interested in current efforts which employ applicable
data discovery and mining approaches to Web categorization.
All additional projects and studies will be integrated within the
Project Aristotle(sm) site after review.
Regards,
Gerry McKiernan
Curator, CyberStacks(sm)
Iowa State University
152 Parks Library
Ames IA 50011
Previous2NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Wed, 14 Aug 1996 18:30:25 +1000
From: Thierry Van de Merckt (thierry@cs.usyd.edu.au)
Organization: cs.usyd.edu.au
X-Mailer: Mozilla 2.02 (X11; I; SunOS 5.5 sun4d)
MIME-Version: 1.0
To: kdd@gte.com
Subject: RE: PR NewsWire: DCI Data Warehouse World, Data Mining Gold Rush,...
> Subject: PR NewsWire: DCI Data Warehouse World, Data Mining Gold Rush,...
>
> This article comes from Ziff-Davis Personal View, but I could not
> find a direct http pointer. -- GPS
Dear Gregory,
More information on the DCI Data Warehouse World in New York can be found
at DCI's Warehouse World site http://www.dciexpo.com/datawhse/.
However,
I did not find a direct pointer to this paper, even there!
For your information, the last benchmark test undertaken at DCI's data
warehousing conference in Santa Clara, June 11-13, has been won by the
Pilot Discovery Server software, according to Pilot themselves http://www.pilotsw.com/about/pressrel/pr619.htm.
As data mining becomes a hot business niche, it will become more and more
difficult to find one's way in all products, awards, benchmarks, etc.
As a consequence I started to build The Corporate KDD Bookmark http://www.cs.su.oz.au/~thierry/ckdd.html
in order to centralise
information on KDD (with an emphasis on Data Mining). What discriminates
this bookmark from The Data Mine or The Knowledge Discovery Mine is that it
targets Business only: software houses, consultants, service providers and
non-profit associations that group many corporate acting in this field. If
you think that this bookmark can be of any help, feel free to post its
address.
Best regards,
-Thierry
--
Dr. Thierry Van de Merckt
The Basser Dept of Computer Science Voice: +61.2 - 351 4926
Madsen F09, room G84 Fax : +61.2 - 351 3838
University of Sydney email: thierry@cs.usyd.edu.au
NSW 2006 - Australia http://www.cs.su.oz.au/~thierry/
Previous3NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Sender: thierry@cs.usyd.edu.au
Date: Wed, 21 Aug 1996 15:59:59 +1000
From: Thierry Van de Merckt (thierry@cs.usyd.edu.au)
Organization: cs.usyd.edu.au
Subject: Corporate Kdd
Dear Gregory,
just to let you know that the Corporate KDD bookmark http://www.cs.su.oz.au/~thierry/ckdd.html
is
up-to-date for the data mining slot.
It's ready for use: all links are accurate and
operational.
Kind regards,
-Thierry
Previous4NextTop
>~~~Siftware:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
(via neuron-digest)
Subject: Web Application
From: George Cybenko (gvc@dartmouth.edu)
Organization: Dept. of Math. and Comp.Sci.; Leiden Univ.; Leiden; the Netherlands
Date: Wed, 31 Jul 1996 17:08:50 +0200
The Informant
Some of my students and I have developed a Web application called
'The Informant'. Its URL is http://informant.dartmouth.edu.
This application saves key words that you provide and runs
periodic web searches (using Lycos now and other search engines
in the near future as well) on your behalf. When new web sites
are found that match your key words with a high score, you are sent
email notifying you that new material has been found. You then
reconnect to http://informant.dartmouth.edu
and get a customized
page showing the new links. The service is free.
For example, suppose you are interested in reinforcement learning.
You would enter 'reinforcement learning' as a search key word. Then
when new material relevant to 'reinforcement learning' is found on the
web (that you have not already seen), you will get notified via email.
It works like a web clipping service essentially.
Although this is not a neural network application per se, it is the
simplest example of a network 'agent' and has turned out to be very
successful.
There are over 5300 registered users already from around the world
and most feedback has been very positive. The application has
been featured on Netscape's 'What's New' button as well as being
the Microsoft Network 'Pick of the Day'. We welcome new users and
feedback about the application, especially ways in which it can
be improved.
The central element of Kepler's architecture is its extensibility
through a 'plug-in' interface for external tools (cf. KDD96 paper, ftp://ftp.gmd.de/ml-archive/GMD/papers/ML75.ps.gz.
This external tool
interface allows new algorithms for existing or new analysis tasks to
be rapidly plugged into the system without redeveloping the system
core. Currently available are e.g. plug-in algorithms for decision
tree induction, backprop neural networks, pattern detection,
clustering, nearest neighbor learning, multiple adaptive regression
splines, and first-order learning tools.
Kepler embeds its analysis functionality in a graphical environment
based on the elementary concepts of spreadsheets and relational
databases. Data are stored in one or more tables and can easily be
imported from a variety of formats. Several functions are offered for
viewing, manipulating, selecting and transforming the available data.
(e.g. for aggregation or sampling). User-specific import formats and
preprocessing operations can be added as plug-ins. The knowledge
generated in an analysis can be graphically visualized; any analysis
run can be inspected, modified, and repeated at any time.
*Discovery tasks: Classification, Clustering, Regression, Subgroup Discovery,
Visualization (extensible through plug-ins)
*Comments:
*Platform(s): Windows 95, Unix
*Contact: Dr. Stefan Wrobel
GMD, FIT, Schloss Birlinghoven, 53574 Sankt Augustin, Germany
stefan.wrobel@gmd.de
*Status: Research Prototype
*Source of information: developer
*Updated: 1996-08-22 by S. Wrobel, e-mail
============================================================================
--------------------------------------------------------------
Dr. Stefan Wrobel
GMD -- German Natl. Research Center for Information Technology
FIT.KI, Schloss Birlinghoven, 53754 Sankt Augustin, Germany
Tel.: +49/2241/14-0, Fax: -2889 E-Mail: stefan.wrobel@gmd.de
WWW http://nathan.gmd.de/persons/stefan.wrobel.html
Secr.: D. Boethgen Tel. -2731, E-Mail: dagmar.boethgen@gmd.de
--------------------------------------------------------------
Previous6NextTop
>~~~Meetings:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
X-Organisation: Social Science Informatics, University of Amsterdam
X-Address: Roetersstraat 15, 1018 WB Amsterdam, The Netherlands.
X-Phone: +31 20 5256789
X-Fax: +31 20 5256896
Date: Mon, 19 Aug 1996 13:31:53 +0200
From: Maarten van Someren (maarten@swi.psy.uva.nl)
To: ilpnet@ijs.si,
inductive@hermes.csd.unb.ca,
kdd@gte.com,
ml@ics.uci.edu,
mlnet@swi.psy.uva.nl,
schlimme@eecs.wsu.edu
Subject: ECML-MLnet workshops 1997
Cc: RADIM@vse.cz,
aha@AIC.NRL.Navy.Mil,
gerhard@ai.univie.ac.at,
kaiser@i60s36.ira.uka.de,
maarten@swi.psy.uva.nl,
malerba@lacam.uniba.it,
mantaras@iiia.csic.es,
saitta@di.unito.it
Content-Length: 2179
9th European Conference on Machine Learning (ECML'97)
ECML-MLNet Workshops
Call for Workshop Proposals
Joint with the European Conference on Machine Learning 1997 workshops
will be organised. These workshops will be organised in collaboration with
the European Network of Excellence in Machine Learning MLnet.
The workshops will be held on 26 April 1997, directly after
ECML-97. One important goal for workshops is to reinforce the relations
between Machine Learning and other fields such as robotics, knowledge-based
systems, natural language processing, neural nets, statistics, human-computer
interaction, human learning or databases. Workshops can use the conference
facilities for publicity and meeting rooms. The format and programme of a
workshop are responsibility of the organisers. The organizing committee
should include at least one member associated with an MLnet node. There will
be a possibility for MLnet members to apply for travel grants to attend the
workshops. Proceedings can be produced in the form of a technical report in
collaboration with the ECML-97 organisation. The schedule is the same for all
workshops and is given below.
----------------------------------------------
Workshop proposals will be evaluated by: relevance to the field of
machine learning, added value with respect to the conference programme
and estimated attendance.
Proposals
Workshop proposals must contain the following information:
* Name, address and email of a contact person
* Persons who are responsible for the programme
* Draft call for contributions
* Indication how the workshop complements the ECML conference
* Description of the intended participants and audience of the workshop
Send proposals for ECML workshops in hardcopy and by email to:
Maarten van Someren
Social Science Informatics
Faculty of Psychology
University of Amsterdam
Roetersstraat 15
1018 WB Amsterdam
email: maarten@swi.psy.uva.nl
Schedule:
Deadline for workshop proposals: 21 October 1996
Acceptance notification: 4 November 1996
Paper submissions: 15 February 1997
Acceptance: 8 March 1997
Camera ready copy: 1 April 1997
Previous7NextTop
>~~~Meetings:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Tue, 20 Aug 1996 10:54:14 +0900
From: tsumoto.com@mri.tmd.ac.jp
(Shusaku Tsumoto)
Subject: Web Page for RSFD'96
Dear Rough Setters,
On behalf of the Organizing Committee of RSFD'96,
I would like to inform all of you that
a Web Page for the Fourth International Workshop on Rough Sets,
Fuzzy Sets and Machine Discovery (RSFD'96) is now available.
You can access Call for Participation and Camera Ready Instructions
from this page, and in the near future,
all the information on RSFD'96 will be accessible from this Web Page.
Sincerely yours,
Shusaku Tsumoto
+----------------------------------------------------------------------+
| |& %| | Shusaku Tsumoto |
| | & % | |~~| | Assistant Professor |
| | V | |~~|& | Department of Information Medicine |
| Medical Research Institute |
| TEL: +81-3-5803-5840 Tokyo Medical and Dental Unversity |
| FAX: +81-3-5803-0247 1-5-45 Yushima, Bunkyo-city, |
| E-mail: tsumoto.com@mri.tmd.ac.jp
Tokyo 113 Japan |
| << Forever Voyaging ....>> |
+----------------------------------------------------------------------+
Previous8NextTop
>~~~Meetings:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: Matthias Klusch (mkl@informatik.uni-kiel.de)
Subject: Final CFP: Workshop CIA-97 on Copperative Information Agents
Date: Fri, 16 Aug 96 12:14:09 METDST
FINAL CALL FOR PAPERS
*****************************************************************************
1. International Workshop CIA-97
COOPERATIVE INFORMATION AGENTS - DAI meets Database Systems
26th (Wed) - 28th (Fri) of February 1997
University of Kiel, Computer Science Department,
Kiel, Germany
*****************************************************************************
The workshop CIA-97 will be held in cooperation with the research groups on
- Distributed Artificial Intelligence (DAI) FG 1.1.6,
- Database Systems FG 2.5.1, and
- Methods for Information Systems Development (EMISA) FG 2.5.2
of the German Society for Computer Science GI.
DESCRIPTION and TOPICS:
-----------------------
This workshop will focus on issues concerning approaches for an integrated
use of methodologies from both research areas, DAI and Database Systems,
especially for the development of cooperative information agents.
The idea of a partial synthesis of both, the DAI as well as the Database
research area seems to be very promising. One attempt to capture possible
benefits from such a synthesis has lead to the introduction of the paradigm
of Cooperative Information Systems (CIS) in 1992.
In our context a CIS is constituted by a set of intelligent agents where
each of them is uniquely attached to one database system. Such information
agents behave like active, intelligent database front-ends trying to satisfy
their own application-specific task goals alone or in utilitarian cooperation
with others. In particular the necessity to respect the database autonomy
requirements hinders such cooperation e.g. for information gathering.
In addition, as the Internet becomes increasingly commercialized,
the information agents will receive monetary rewards for their services and
may negotiate with each other to maximize their expected utility.
The design of information agents inherently requires knowledge from several
different research areas like DAI, Database and Expert Systems, and AI.
Unfortunately, practical and theoretical work which is relevant for the
development of information agents tends to be scattered across several
different forums of respective computing subareas: There is an obvious need
for a survey of these works, their advantages and limitations.
Thus, the workshop aims for being a small but intensive forum for a presentation
and exchange of ideas, work in progress, reviews as well as an engaged discussion
between all attendees.
Topics of interest include, but are not limited to:
o architecture of information agents,
o autonomy requirements of information systems and their impacts
for the development of cooperative information agents,
o decentralized construction and management of common ontologies
for cooperative information agents,
o knowledge discovery and data mining for information evolution
in large database networks,
o semantic querying in multidatabase systems,
o use of object-oriented modelling within the design of information agents,
o methods for utilitarian coalition formation among autonomous agents,
o impacts and handling of lying agents for cooperative information search,
o approaches towards a theory of organization in multi-agent systems,
o adaptation and self-organization of information agents
in changing environments,
o planning in cooperative search for information,
o user interface issues for information agents,
o evaluation and development environments for information agents,
o agent communication languages for information agents,
o approaches towards mobile information agents in the Internet
(e.g. Telescript, Java) ,
o interconnection of autonomous and heterogeneous databases in the WWW,
o security aspects for information agents.
CONTRIBUTIONS:
--------------
Authors are invited to submit papers describing both theoretical and practical
work dealing with the use of methods from distributed artificial intelligence
for cooperation between a set of heterogeneous, autonomous databases.
Papers which describe ongoing research or provide an excellent surveying work
are in particular welcome.
Topics of interest include, but are not limited to the ones listed above.
Format:
^^^^^^^
The paper must be formatted in A4 size using 10 point Times.
Printing area should be 12.2 x 19.3 cm.
Each submission must have in addition a separate title page which includes the title,
the 300-400 word abstract, a list of keywords, the names and addresses of all authors,
their email addresses, and their telephone and fax numbers.
For publication in the Springer series LECTURE NOTES IN ARTIFICIAL INTELLIGENCE
please prepare a camera-ready version of your contribution
following the guidelines and LNCS latex style file available at ftp://hugo.informatik.uni-kiel.de/pub/mkl/cia97/lnai.txt
and ftp://hugo.informatik.uni-kiel.de/pub/mkl/cia97/llncs.sty
or send a request by EMail to lncs@springer.de
(for their 'LNAI Advice to authors for the preparation of camera-ready contributions
to LNAI Proceedings') and send a request by EMail to svserv vax.ntp.springer.de
containing the line 'get /tex/latex/llncs.zip' for LaTeX or 'get /tex/plain/plncs.zip'
for plain TeX.
The length of submitted camera-ready papers must be no more than 12 pages including
all figures, tables, and bibliography. All papers must be written in English and neither
accepted nor under review by other conferences or journals. Papers not conforming to the
above requirements may be rejected without review. Papers can be submitted both by mail
and electronic mail. Electronic submission must be in postscript format.
Submission:
^^^^^^^^^^^
Send three hard-copies or the postscript file of your contribution to
Matthias Klusch
Institut f'ur Informatik und Praktische Mathematik
Christian-Albrechts-Universit'at zu Kiel,
Olshausenstr. 40, 24118 Kiel
EMail: mkl@informatik.uni-kiel.d400.de
It is recommended to send your contribution as a postscript file by email.
In order to inform us about your submission, please send an email to:
{ak,mkl}@informatik.uni-kiel.d400.de
Proceedings:
^^^^^^^^^^^^
The workshop proceedings including all accepted papers as well as the
invited papers will be available for all registered participants at the workshop.
The proceedings will be published in the LNAI Series (Springer Verlag).
IMPORTANT DATES:
----------------
Camera-Ready Paper Submission Deadline : 3rd of November 1996
Notification to the Authors : 15th of December 1996
PARTICIPATION:
--------------
The number of participants will be restricted to atmost 80 peoples.
Preference is given to people with accepted contributions.
The registration fee amounts to 80,-DM.
Further informations on registration concerning the method
of fee payment and the registration form
are available on the CIA-97 WWW page http://www.informatik.uni-kiel.de/~mkl/registration.html
and will be given also in forthcoming call for participation.
Full registration must be done before 10th of January 1997.
For pre-registration please send your name, full address (incl. phone/fax),
affiliation, and a short description about your current research interests to the
Organization Committee (see below). Let us also know if you intend to submit a paper.
DATE AND LOCATION:
------------------
The workshop starts on Wednesday, 26th of February 1997, at 8am and
ends on Friday, 28th of February 1997, afternoon.
It takes place at the University of Kiel
(Institut f'ur Informatik, Olshausenstr. 40, 24118 Kiel, Germany).
Please look at the WWW-page http://www.informatik.uni-kiel.de/~mkl/cia97.html
of the workshop for more information about location, registration, travel
and accomodation.
PROGRAM COMMITTEE:
Wolfgang Benn (University of Chemnitz, Germany)
Sonia Bergamaschi (Universita' di Modena, Italy)
Hans-Dieter Burkhard (Humboldt University Berlin, Germany)
Misbah Deen (University of Keele, UK)
Yves Demazeau (Leibniz/Imag/CNRS, France)
Frank Dignum (University of Eindhoven, The Netherlands)
Edmund Durfee (University of Michigan, USA)
Tim Finin (University of Maryland, USA)
Klaus Fischer (DFKI Saarbruecken, Germany)
Joachim Hammer (Stanford University, USA)
Peter Kandzia (University of Kiel, Germany)
Larry Kerschberg (George Mason University, USA)
Stefan Kirn (University of M'unster, Germany)
Matthias Klusch (University of Kiel, Germany)
Sarit Kraus (Bar Ilan University, Israel)
Klaus Meyer-Wegener (University of Dresden, Germany)
Joerg P. Mueller (DFKI Saarbruecken, Germany)
Aris Ouksel (University of Illinois, USA)
Mike P. Papazoglou (QUT Brisbane, Australia)
Jeffrey Rosenschein (Hebrew University, Israel)
Tuomas Sandholm (University of Massachusetts at Amherst, USA)
Onn Shehory (Bar Ilan University, Israel)
Antonio Si (Polytechnic University, Hong Kong)
Gottfried Vossen (University of Muenster, Germany)
Gerd Wagner (University of Leipzig, Germany)
Mike Wooldridge (Manchester Mteropolitan University, UK)
General Chairs and Organization Committee:
Matthias Klusch (University of Kiel, Germany)
Peter Kandzia (University of Kiel, Germany)
KDD Nuggets Index