(text)
Hiroshi Motoda, Call for Paper of PKAW98,
Singapore, November 22-23, 1998 http://www.ar.sanken.osaka-u.ac.jp/PKAW98.html
--
on the latest news, publications, tools, meetings, and other relevant items
in the Data Mining and Knowledge Discovery field.
KD Nuggets is currently reaching over 5000 readers in 70+ countries
2-3 times a month.
Items relevant to data mining and knowledge discovery are welcome
and should be emailed to gps
in ASCII text or HTML format.
An item should have a subject line which clearly describes
what is it about to KDNuggets readers.
Please keep calls for papers and meeting announcements
short (50 lines or less of up to 80-characters), and provide a web site for
details, such as papers submission guidelines.
All items may be edited for size.
Back issues of KD Nuggets, a catalog of data mining tools
('Siftware'), pointers to data mining companies, relevant websites,
meetings, etc are available at KDNuggets Directory at http://www.kdnuggets.com/
********************* Official disclaimer ***************************
All opinions expressed herein are those of the contributors and not
necessarily of their respective employers (or of KD Nuggets)
*********************************************************************
~~~ Quotable Quote:
3. How long is this Beta guy going to keep testing our stuff?
actual quote from a Dilbert quotes contest, run by a magazine
(thanks to Kathleen Wright) Previous1NextTop
Date: Thu, 11 June 1998 09:41:10 -0500 (EST)
From: Gregory Piatetsky-Shapiro gps
Subject: KDD-98 Preliminary Schedule available
Web: www.kdnuggets.com/meetings/kdd98
KDD-98 Conference (New York, NY, Aug 27-31, 1998)
preliminary schedule is available at
Internet Week (June 8, 1998) has an interesting article by David Carr,
'Data Mining Makes Slow Transition to Internet'
The article talks about net data mining at New York Times, Microsoft,
Amazon.com, U.S. Military and more.
Another article on Data Mining (thanks to Susan Tafolla)
written by
Christopher Elliott, appeared June 1, 1998 edition of InternetWeek.
Data Mining: Give Your Data a Workout
When it comes to data mining, it's easy to feel like you're on a
treadmill. Sure,computers can process, analyze, slice and dice data faster
and more precisely than ever.
And the Web does offer an easy-to-use and inexpensive
dissemination tool. But if the information you're mining is out of shape,
you're just wasting calories. It's up to you and your IT staff to give
your data the workout it deserves. Only then can data mining live up to
its promise.
Previous3NextTop
Date: Mon, 08 Jun 1998 14:55:12 +0100
From: Joe Evans joee@isl.co.uk
Subject: Data mining process model published and piloted
FOR IMMEDIATE RELEASE
Contact: Joe Evans
(01256) 355 899 joee@isl.co.uk
1st June 1998
Data mining process model published and piloted
Second CRISP-DM Workshop Held
London, May 1998 - More than 20 key players in the data mining market
have met to discuss the first draft of a new process model, CRISP-DM.
The model is designed to help businesses plan and work through the
complete data mining process, from problem specification to deployment
of results.
The CRISP-DM initiative - CRoss-Industry Standard Process for Data
Mining - is partially funded by the European Commission. The core
consortium consists of NCR, ISL, Daimler-Benz and OHRA.
The first draft of CRISP-DM has already been piloted in applications at
Mercedes-Benz and OHRA. Initial results are said to be very positive.
At the centre of the CRISP-DM project is a Special Interest Group (SIG)
of data mining tool and service suppliers, together with large-scale
commercial users. The SIG continues to grow, and there are now more than
80 members world-wide.
This, the second SIG workshop, built on the presentations and
discussions from the first workshop held in Amsterdam last November. SIG
members, including IBM and Data Distilleries, provided input from
project experience and feedback on the draft process model.
The CRISP-DM partners presented the latest developments to the process
model, based on experience from commercial data mining projects using
the initial process model, and on input from the SIG members. They also
demonstrated integrated software support for the process, in ISL's
Clementine Data Mining System.
'The changes we made have been welcomed by others in the field who will
play a vital part in establishing CRISP-DM as the de facto standard for
data mining,' said Jens Hejlesen of NCR, CRISP-DM project manager.
'CRISP-DM has an important role in allowing companies to leverage
investment in data warehousing through data mining.'
Members are still being recruited for the CRISP-DM SIG, and further
workshops are planned during the next few months.
The finalised CRISP-DM model is due to be
published by the end of the year.
Hi,
I was wondering if anybody knows where I can find some sample datasets of
retail purchases. For example, a snapshot of all sales receipts for a
particular store location (or department) over a three month period.
There is an interesting article in The Scientist about the ARROWSMITH
program for literature-based discovery which was described in the
journal Artificial Intelligence as part of the Vol1/No2 1997 special
issue on Scientific Discovery. The article can be read at:
Data mining is the automatic discovery of patterns, changes,
associations and anomalies in large data sets. Data mining is emerging
as a key enabling technology for a variety of scientific, engineering,
medical and business applications. This special issue of Data Mining
and Knowledge Discovery addresses the following key issues:
Scaling data mining algorithms, applications and systems to massive
data sets. Today's data mining tools are able to deal with moderate
amounts of data, in the range of several million data items. Data
mining over large data sets can take a prohibitive amount of time due
to the computational complexity of the algorithms. The special issue
will highlight techniques from high performance and parallel computing
and their applications to data mining.
Developing data mining algorithms, applications and systems for
mining distributed data. Large-scale data sets are usually logically
and physically distributed, and organisations that are geographically
distributed need a decentralised approach to decision support.
Therefore the issues concerning modern organisations are not just the
size of the data to be mined, but also its distributed nature. The
special issue will highlight distributed data mining and distributed
data intensive decision support.
Integrating data mining with other systems and applications. The
goal of many data mining applications is to derive timely advantageous
knowledge from the data sources available. A major challenge for the
data mining community is not only to develop data mining applications
but also to integrate them effectively with other applications,
systems, and business processes throughout a large scale
enterprise. The special issue will highlight techniques of integrating
data mining with other application in a distributed computing
environment.
Hello Gregory,
Some of the people at MITRE Corp. suggested I contact you.
Fast Fuzzy Cluster enables real-time clustering of gigabyte size data spaces.
FFC uses no distance metric and is massively parallel.
Please check out my web page at http://members.aol.com/awareai.
I hope you are interested and will include it in nuggets.
Thank you.
Michael Bickel AwareAI@aol.com
Implementing Data Mining and Knowledge Discovery
-A Course Developed by Intervista Strategic Development
Toronto, Ontario (Canada)
June 15-16/98
Data Mining can provide organizations with the strategic advantage
needed for survival and growth in today's competitive
environment. This educational session is designed to provide
participants with an opportunity to understand when to use
Data-mining, and how to make it work effectively for you and your
business. Implementing Data Mining and Knowledge Discovery is the
perfect opportunity to get answers to your important technology and
management questions, while providing you with the much needed insight
into choosing the proper and successful tools for Data mining
implementation and for your Data and knowledge needs.
For more information on the course, registration and prices visit our
website at: http://www.cam.org/~ivista/datamining.html
or call
1-800-397-9744 Early enrolment and group rates available!
Developed by Intervista Strategic Development Leaders in Knowledge
Management Courses including the latest in Data Warehousing,
Enterprise Document Management, Enterprise Architecture, Data Mining &
More!
We're looking to get a good cross section of users and developers
of commercial data mining software for this workshop. I encourage
readers of the KDD list to submit a position paper and participate
in the workshop!
- kurt
Second Call for Participation
Workshop: Keys to the Commercial Success of Data Mining
To be held in conjunction with The Fourth International Conference on
Knowledge Discovery and Data Mining
New York City, August 31, 1998 http://www.aaai.org/Conferences/KDD/1998/
Chairs:
Kurt Thearling
Director of Analytics
Exchange Applications
One Lincoln Plaza
Boston, MA 02111
Roger M. Stein
Vice President, Senior Credit Officer
Quantitative Analytics and Knowledge Based Systems
Moody's Investors Service
99 Church Street
New York, NY 10007
Data mining is on the cusp of true commercial success. Commercial
institutions are starting to move beyond pilot studies and research programs
toward the production use of predictive models for real world business
applications. While this is exciting, it is also where it gets harder.
Successful data mining in business doesn't come down to simply having a hot
algorithm and giving it to an experienced modeler. Business users care
about things such as database support, application integration, business
templates, flexibility, scalability, real profitability, and other issues
that have not historically been the concern of the KDD community.
From a development point of view, the core algorithms are now a small part,
perhaps 10%, of the overall data mining application, which itself is only
10% of the business process that contains the application. The purpose of
this workshop is to focus on the remaining 99% so that commercial data
mining application are relevant to business users.
The goal is to bring together a diverse group of developers, users, and
integrators of business data mining applications. The workshop will
consist of a number of in-depth case studies and analyses, several invited
speakers, and panel sessions. Time will also be set aside for
discussions.
It is expected that the workshop will include forty to fifty participants.
Approximately half of the participants will come from the data mining
development community with the other half coming from the data mining
business user community. Developers of commercial software for data
mining will also be eligible to attend the workshop if they have significant
contributions to make beyond promotional pitches. The set of business
users attending will be selected from a diverse set of industries such as
banking, retail, insurance, government, internet services, telecom, etc. In
addition to developers and users, a small number of participants will come
from system integration and services companies.
Position Paper Submission:
All participants must submit a position statement (about two pages)
describing their views on the subject of commercial data mining.
The focus should be on the practical application of data mining
rather than the underlying algorithms.
For business users, position paper topics might include:
- Experiences (positive or negative) regarding the use of data
mining software for commercial applications;
- Areas needing improvement in data mining software;
- Issues in problem formulation for business domains;
- The impact of data mining applications on business processes;
- The lifecycle of mining projects in commercial organizations; or
- Potential data mining applications in specific business domains.
For developers, examples of some possible topics include:
- Issues of database integration for data mining;
- Automated model selection;
- Strategies for addressing data problems;
- Integration with other business software applications;
- Issues in the design of business templates; or
- User interface design for business datamining.
In addition to the position statements, participants need to include the
following information in their submission:
1) Developers: Software developers should include the name of their
software application, technical specifications, and the names of
three representative customers with deployed applications.
2) Users (and SI/Services): End users should include the names of
data mining applications that they have worked with, the industry
that they are working in, and the general problem space they are applying
data mining to.
Participants will be chosen based on position statements and their ability
to contribute to the workshop. All position statements will be
distributed to each attendee before the workshop. Depending on the content
and variety of submissions, a collection of the papers from the workshop may
be published either in book form or as a special issue of a relevant
journal.
Cost: $100 (includes proceedings and lunch)
Timetable:
Jun 15: Papers due
Jul 10: Notification of acceptance/rejection
Aug 31: Workshop
Call for Papers and Participation
Workshop on Knowledge Discovery and Data Mining
5th Pacific Rim International Conference on Artificial Intelligence
(PRICAI-98)
Knowledge discovery and data mining (KDD) has become an active and
growing research area. It is not only of academic interest, but also of
great practical significance. It has attracted a large number of researchers
and practitioners from many disciplines, e.g., machine learning, databases,
AI, statistics, and data visualization. The reason for this tremendous
interest in KDD is obvious. Due to the rapid computerization of the past
two decades, almost all organizations and companies have collected a huge
amount of data in their databases. These organizations and companies need
to understand their data and/or to discover useful knowledge from the data
that can be used for a competitive advantage. KDD aims to help them to do
just that.
This workshop aim to bring together researchers and practitioners from
various disciplines concerned with mining or discovering useful knowledge
from data. The objective of this meeting is to discuss the following issues:
What are the major challenges in data mining applications?
What are promising research directions in solving these challenging problems?
What are the most promising directions for cross-disciplinary research?
Important Dates
Papers due by: AUG 25, 1998
Notification of Acceptance: Sep 25, 1998
Camera-ready version of Final Paper due: Oct 25, 1998
Date of Workshop: Nov 23
Main PRICAI-98 Conference: Nov. 22 - 27, 1998
(Check the workshop web page for further information)
Leonard N. Stern School of Business
New York University
Computational Finance (CF99)
January 6, 1999 (Tutorials)
January 7 - 8 (Conference)
The sixth international conference Computational Finance (CF99) will
be held at NYU's Leonard N. Stern School of Business. CF99 is
sponsored by the New York University Salomon Center, the Center for
Research on Information Systems and the Department of Statistics and
Operations Research.
Computational Finance has emerged as a genuinely cross- disciplinary
research meeting. CF99 is the sixth in a series of conferences that
have been sponsored by the California Institute of Technology and the
London Business School. In the past, this conference was called Neural
Networks in the Capital Markets (NNCM). The expanding set of
computational tools has moved this meeting from its original emphasis
on neural network techniques to a broad spectrum of different methodologies.
With several hundred attendees, this fully refereed conference has
become an international forum where original research in advanced
computational applications in finance is presented and discussed. CF99
brings together decision-makers and strategists from the financial
industries, with academics from finance, statistics, economics,
information systems and other disciplines. In the last few years, the
conference has seen papers covering many different computational
techniques including: statistical machine learning, Monte Carlo
simulation, data mining, knowledge discovery, bootstrapping, genetic
algorithms, nonparametric methods, information theory and fuzzy logic.
Applications in many different areas are welcome, including but not
limited to: risk management, asset allocation, dynamic trading and
hedging strategies, forecasting, numerical solutions of derivative
PDEs, exotic options and trading cost control.
Studies may cover any major international financial market including
equity, foreign exchange, bond, commodity and derivatives. The
conference emphasizes in-depth analysis and comparative evaluation
with established approaches.
CF99 begins with a full day of tutorials designed to inform the
diverse group of participants on a selection of the latest tools and
research results. Tutorial speakers include Professor Stephen
Figlewski of the Stern School of Business. The conference also
features several invited speakers sharing their expertise from both
the academic and applied perspectives. The keynote speaker is David
E. Shaw, PhD, Chairman and CEO of D. E. Shaw & Co., Inc.
The conference will have several talk and poster sessions for accepted
papers. A selection of the presentations will be invited to appear in
a volume published by Kluwer Academic Publishers.
Submissions to CF99:
Authors who wish to present papers should submit four copies
along with full contact information, including e-mail addresses, to:
CF99 / Andreas Weigend
Information Systems Department
Leonard N. Stern School of Business
New York University
44 W 4th St., MEC 9-171
New York, NY 10012, USA
All submissions must be received by August 15, 1998. Full papers are
preferred, but extended abstracts clearly stating the results are
acceptable. Only original, relevant research work will be accepted.
Call for Participation
JICSLP'98 Post-Conference Workshop
CompulogNet Area Meeting on
Computational Logic and Machine Learning
June 20th, 1998
Manchester, UK
Organiser: John Lloyd, University of Bristol
The next CompulogNet Area Meeting on 'Computational Logic and
Machine Learning' will be held as a Post-Conference Workshop
at JICSLP'98. This meeting is sponsored by the ESPRIT Network of
Excellence in Computational Logic (CompulogNet).
Highlights of the meeting:
1. Two invited overview talks by ML experts.
2. A panel discussion on the big issues in inductive learning.
3. An interesting programme of submitted papers.
The theme of the meeting will be
'Logic Programming and Machine Learning: A Two-way Connection'.
There is a two-way connection between logic programming and machine
learning. For example, LP has already significantly influenced
(symbolic) ML through the field of inductive logic programming.
There is potential for even greater influence in the near future,
for example, through the application of constraint or higher-order
LP languages, and through the use of abduction. On the other hand,
ML has influenced LP by providing an application area full of
industrially significant problems which can provide a challenge
for the most sophisticated and up-to-date techniques of logic
programming.
To make the meeting attractive to logic programmers who know little
about machine learning, the meeting will start with two invited
overview talks by experts in machine learning.
The meeting will end with a panel discussion on inductive learning
which is intended to highlight the likely major research issues in,
and applications of, inductive learning over the next 5 to 10 years.
This meeting will be an excellent opportunity for logic programmers
to learn about an exciting application area for computational logic
and also for machine learners to find out about recent advances in
computational logic which have applications to machine learning.
For details on the programme, follow the link to the workshop from:
SECOND CALL FOR PAPERS: PAKDD-99
The Third Pacific-Asia Conference on
Knowledge Discovery and Data Mining
-----------------------------------
Xiangshan Hotel, Beijing, China
===============================
April 26-28, 1999
Sponsored by:
Tsinghua University
National Science Foundation of China
Chinese Computer Federation
Toshiba Corporation
NEC Software Chugoku, Ltd.
Invited Speakers:
Won Kim (Keynote speech, Cyber Database Solutions, USA)
Hiroshi Motoda (Osaka University, Japan)
The Third Pacific-Asia Conference on Knowledge Discovery and Data
Mining (PAKDD-99) will provide an international forum for the sharing
of original research results and practical development experiences
among researchers and application developers from different KDD
related areas such as machine learning, databases, statistics,
knowledge acquisition, data visualization, knowledge-based systems,
soft computing, and high performance computing. It will follow the
success of PAKDD-97 held in Singapore in 1997 and PAKDD-98 held in
Australia in 1998 by bringing together participants from universities,
industry and government.
Call for Papers
PKAW98, The 1998 Pacific Rim Knowledge Acquisition Workshop
Sponsored by PRICAI98
Venue & Date
Singapore, November 22-23, 1998
1. Introduction
The objective of this workshop is to assemble theoreticians and
practitioners concerned with developing methods and systems that
assist the knowledge acquisition process and assessing the suitability
of such methods. Thus, the workshop includes all aspects of
eliciting, acquiring, modeling and managing knowledge, and their role
in the construction of knowledge-intensive systems. Knowledge
acquisition still remains the bottleneck for building a knowledge based
system. Reuse and sharing of knowledge bases are major issues and
no satisfactory solutions have been agreed upon yet. There is a wide
range of research. Much of the work in this field has been knowledge
acquisition from human experts. The advent of the age of digital
information has brought the problem of data overload. Our ability to
analyze and understand massive datasets lags far behind our ability to
gather and store the data. A new generation of computational
techniques and tools is required to support the acquisition of useful
knowledge from the rapidly growing volume of data. All of these are to
be discussed in this workshop.
This workshop offers an opportunity to draw together both aspects of
dealing with the situated nature of human knowledge and expertise and
of developing methods that depend more on their algorithmic adequacy
than on the expertise of the knowledge engineer.
For details see
2. Topics of Interest
Papers are invited in all aspects of knowledge acquisition for
knowledge-based systems, including (but not restricted to):
o Fundamental views on knowledge that affect the knowledge
acquisition process and the use of knowledge in knowledge
engineering
o Algorithmic approaches to knowledge acquisition
o Tools and techniques for knowledge acquisition, knowledge
maintenance and knowledge validation
o Evaluation of knowledge acquisition techniques, tools and methods
o Knowledge acquisition, machine learning and knowledge discovery
o Languages and frameworks for knowledge and knowledge modeling
o Integration of knowledge acquisition techniques with wider
information systems or decision support systems
o Methods and techniques for sharing and reusing knowledge
o Distributed knowledge acquisition through infrastructures such as
the Internet
3. Important Dates
Papers due by: July 10, 1998
Notification of Acceptance: September 10, 1998
Camera-ready version of Final Paper due: October 10, 1998
Date of Workshop: November 22-23, 1998