--
Data Mining and Knowledge Discovery community, focusing on the latest
research and applications. Submissions should be emailed,
with a DESCRIPTIVE subject line (and a URL, when available) to kdd@gte.com
To subscribe, email to kdd-request@gte.com
message with
subscribe kdd-nuggets
in the first line (the rest of the message and subject are ignored).
See http://info.gte.com/~kdd/subscribe.html
for details.
Nuggets frequency is approximately 3 times a month.
Back issues of KD 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 http://info.gte.com/~kdd/
-- Gregory Piatetsky-Shapiro (editor)
(p.s. this is my last week at GTE.
Starting today, I can be reached at gps
.
After March 1, 1997 I will continue to edit and distribute KD Nuggets
and maintain KD Mine pages at a new web site -- details to be announced soon!
The kdd@gte.com
and kdd-request@gte.com
email addresses would still
work for a while. GPS)
********************* 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 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Q: How does the large number of meetings correlate with the
large number of job announcements?
A: Somebody got to work, while all those other people go to meetings Previous1NextTop
>~~~Publications:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Sun, 16 Feb 1997 12:20:06 -0500
From: gps0 (Gregory Piatetsky-Shapiro)
Subject: NeuroVe$t journal and Data Mining for Financial Applications]
Content-Length: 3383
Here, reprinted with permission, is the review of AKDDM book from
***
NeuroVe$t Journal, Jan/Feb 1996, pg.49, Reviews in Brief section -
Advances in Knowledge Discovery and Data Mining
Advances in Knowledge Discovery and Data Mining (AKDDM) provides
a well-edited collection of material from the 1994 KDD (Knowledge
Discovery in Databases) Workshop, and several additional invited papers.
In all, 23 papers presented in 7 chapters are included along with a
useful appendix on KDD terminology and resources on the Internet.
Coupled with an extensive index and a very good job of editing, AKDDM
makes for a very accessible and worthwhile collection of papers.
Of particular interest to investors and traders, especially those
using data-driven computer technologies, are 'A Statistical Perspective
on Knowledge Discovery in Databases' by Elder and Pregibon, which
provides a very good introduction to the topics. 'Finding Patterns in
Time Series' by Berndt and Clifford include in their studies a look at
various technical analysis patterns of daily DJIA prices from 1989 to
1993, using pattern templates that vary in length from 9 to 12 trading
days. 'Integrating Inductive and Deductive Reasoning for Data Mining' by
Simoudis, Livezey and Kerber involves the creation of portfolios of 100
stocks from 7 years of data on 1500 stocks. 'Predicting Equity Returns
from Securities Data with Minimal Rule Generation' by Apte and Hong
describes a minimal rule generation technique for forecasting 1-month S&P
500 returns using 40 fundamental and technical variables (not
specifically identified).
Unfortunately, there is scant mention of the specifics of rough
sets, nearest neighbor classifiers, learning vector quantizers,
self-organizing maps, fuzzy logic and other tools of interest to
practitioners and applied researchers working in the field. And, on more
than a couple of occasions, the authors (including the editors) appear to
venture beyond their respective areas of expertise. However, the few
shortcomings are overshadowed by several very good introductory studies.
Seldom do I recommend collections of workshop or conference papers
to the general audience. However, AKDDM represents an exception.
Despite its weaknesses, it provides a valuable introduction to a
relatively new, yet increasingly important area of applied research.
Financial practitioners who are particularly interested in data mining
will certainly want to take a look.
Usama M. Fayyad, Gregory Piatetsky-Shapiro, Padhraic Smyth, and
Ramasamy Uthursusamy (editors). 1996. The MIT Press, 55 Hayward Street,
Cambridge, MA 02142. 620 pages. US$50. ISBN 0-262-56097-6.
617-253-5643. -- James Hampton
***
(c) Copyright 1997 Finance & Technology Publishing,
P.O. Box 764, Haymarket, VA 20168. Reprinted
with permission of the publisher from NeuroVe$t Journal, Jan/Feb 1997.
Previous2NextTop
>~~~Siftware:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Sat, 15 Feb 1997 12:14:00 -0800
From: Ronny Kohavi (ronnyk@starry.engr.sgi.com)
Subject: SGI MineSet Available for Varsity Members
Reply-to: ronnyk@relay.engr.SGI.COM
Silicon Graphics' MineSet
Available to Varsity Members
----------------------------
MineSet(TM) version 1.1 is the second release of SGI's product for
data mining and exploratory data analysis. MineSet integrates tools
for data access, transformations, analytical data mining, and visual
data mining. See http://www.sgi.com/Products/software/MineSet
for
more information.
In addition to 30-day free evaluation copies available to any site,
with the new release of SGI's Varsity program CDs (happening now),
varsity members can get PERMANENT MineSet licenses.
Any educational institution is eligible. To qualify, the institution
must have an infrastructure capable of handling technical software
support for its Silicon Graphics users who have purchased Varsity
Program software packages. THE VARSITY PROGRAM AGREEMENT MUST BE
COMPLETED AND SIGNED BY THE INSTITUTION AND APPROVED BY SILICON
GRAPHICS.
The institution buys the right to distribute Varsity Program Developer
Package right-to-use licenses in multiples of 10 or 25. These licenses
are maintained by purchasing yearly support. Thus, the cost of
ownership is significantly reduced in the second year and beyond.
To get a permanent license, the site administrator can use the VPX
(varsity ID) number to get a license from http://www.sgi.com/Products/license.html
(click the radio
button for varsity).
Previous3NextTop
>~~~Positions:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: 'Gutschow, Todd' (twg@hnc.com)
Subject: Data Mining Research Position at HNC Software Inc.
Date: Wed, 12 Feb 1997 17:51:58 -0800
The Technology Development Group at HNC Software Incorporated has an
opening for a Manager of Data Mining Technology Research. The Technology
Development Group is responsible for the core data analysis, data
mining, and
data modeling technology used in all HNC vertical solution products. The
position
will report to the Vice President of Technology Development and will be
located at HNC's headquarters facility in San Diego, CA.
Duties/Job Description:
Conduct research in to new data mining algorithms in support of
the Database Mining=D2 Marksman and other HNC products. Identify and
coordinate data mining technology projects across all HNC operating
groups.
Monitor the data mining research literature to identify promising new
techniques.
Support product development and marketing activities via customer
presentations, conference talks, and white papers.
Required Qualifications (Experience/Skills):
MS or Ph.D. in computer science, engineering, mathematics or other
hard science (e.g., physics, chemistry, etc.). Five or more years
experience in implementing and evaluating new statistical data analysis, neural
networks, and/or data mining algorithms. Good software development
skills. Experience with modern software development processes and
tools (e.g., C++, Object oriented design, etc.). Strong communication and
presentation skills.
Preferred Qualifications (Experience/Skills)
Strong algorithm diagnosis and troubleshooting skills. Experience with
database marketing and its associated data analysis problems. Project
management experience.
If you know someone with the above qualifications who is interested in
employment opportunities with HNC, please ask them to fax, mail or
e-mail resumes immediately to:
Human Resources Department
HNC Software Inc.
5930 Cornerstone Court West
San Diego, CA 92121
FAX: (619) 452-6524
E-mail: jobs@hnc.com
Reference Job No. 293 Previous4NextTop
>~~~Positions:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: Colin Shearer (colin@isl.co.uk)
Date: Thu, 13 Feb 97 14:36:13 GMT
Subject: VACANCIES - DATA MINING TOOL DEVELOPMENT & CONSULTING : UK & US
VACANCIES - DATA MINING TOOL DEVELOPMENT & CONSULTING : UK & US
===============================================================
Integral Solutions Limited (ISL) is a leading supplier of advanced decision
support technology, specialising in data mining.
Our award-winning Clementine tool combines multiple modelling techniques
(neural networks, rule induction, regression) with data visualisation and
manipulation to extract high-value decision making knowledge from large bodies
of historical data. A rich visual programming interface makes Clementine
accessible to non-technologist 'data owners' - business, rather than IT,
experts - and provides high productivity for 'power' users. Clementine has
established a leading position in the data mining market, and is in use in a
wide range of industry sectors including finance, retail, telecoms,
pharmaceuticals, utilities, broadcasting, defence. Applications are diverse
and include demand prediction, customer profiling, risk assessment, turnover
forecasting, process optimisation, fault pre-emption and fraud detection.
We have an urgent need to recruit top-quality technical personnel. Current
vacancies are:
Data Mining Tool Developers
---------------------------
Basingstoke, UK.
To work on the ongoing development of Clementine.
Candidates should have an interest in, and ideally experience of implementing,
advanced modelling and data analysis techniques; experience of commercial data
mining tool development is desirable but not essential. Experience of some or
all of the following would also be useful:
Unix GUI Development
VMS Pop11
X Windows / Motif C
Windows 95 / NT SQL
Databases/ODBC Statistics
Applicants should have a 2.1 or better at first degree; a relevant second
degree may be an advantage. Technical excellence is expected, but must be
combined with first rate communications and interpersonal skills and a desire
for close contact with customers. Recent graduates and those with commercial
experience will both be considered.
Data Mining Consultants
-----------------------
Basingstoke, UK; King of Prussia, PA, USA.
To apply Clementine to customers' business problems. The role will include
pre-sales consulting, training, and developing solutions.
Candidates should be degree-qualified (2.1 or better) and, ideally, should
have experience of data analysis and modelling in a business environment.
Excellent communication and interpersonal skills are vital, and candidates
should display initiative, creativity, enthusiasm (and the ability to convey
it to clients) and self-management skills.
As ISL's clients span many markets, our consultants need the ability to
assimilate knowledge of any client's business, understand their problems, and
fit a data mining solution to these. However, we also encourage applications
from those with a specific business/sector specialisation (for example finance
(banking, insurance), retail or manufacturing).
We are willing to consider applications both from experienced consultants and
from any other candidates who believe they have the aptitude to be developed
into first-class consultants.
This is an opportunity to join a small (30 people) but dynamic and rapidly
developing company in an exciting business/technology area. ISL provides a
stimulating and technically challenging environment with considerable scope
for professional development.
ISL is an equal opportunities employer. We encourage applications from new
graduates through to experienced professionals. Salaries/benefits are
competitive, and commensurate with relevant experience.
Please apply with CV to:
For UK: For US:
Linda Montgomery, Kevin Peyton
Integral Solutions Limited, ISL Decsion Systems Inc.
Berk House, 630 Freedom Business Center
Basing View, King of Prussia
Basingstoke, PA 19406
RG21 4RG USA
UK
The Boeing Company, the world's largest aerospace company, is actively
working research projects in advanced computing technologies including
projects involving NASA, FAA, Air Traffic Control, and Global
Positioning as well as airplane and manufacturing research.
The Research and Technology organization located in Bellevue,
Washington, near Seattle, has an open position for a machine learning
researcher. We are the primary computing research organization for
Boeing and have contributed heavily to both short term technology
advances and to long range planning and development.
BACKGROUND REQUIRED: Machine Learning, Knowledge Discovery, Data
Mining, Statistics, Artificial Intelligence or related field.
RESEARCH AREAS: We are developing and applying techniques for data
mining and statistical analyses of diverse types of data, including:
safety incidents, flight data recorders, reliability, maintenance,
manufacturing, and quality assurance data. These are not areas where
most large R&D data mining efforts are currently focused. Research
areas include data models, data mining algorithms, statistics, and
visualization. Issues related to our projects also include pattern
recognition, multidimensional time series, and temporal databases. We
can achieve major practical impacts in the short-term both at Boeing
and in the airline industry, which may result in a safer and more
cost-effective air travel industry.
A Ph.D. in Computer Science or equivalent experience is highly
desirable for the position. We strongly encourage diversity in
backgrounds including both academic and industrial
experiences. Knowledge of machine learning, statistics, and data
mining are important factors. Experience with databases and
programming (C/C++, JAVA, and Splus) is desirable.
APPLICATION: If you meet the requirements and you are interested,
please send your resume via electronic e-mail in plain ASCII format to
zhangw@redwood.rt.cs.boeing.com
(Wei Zhang). You can also send it via
US mail to
Wei Zhang
The Boeing Company
PO Box 3707, MS 7L-66
Seattle, WA 98124-2207
Application deadline is April 30, 1997.
The Boeing Company is an equal opportunity employer.
Previous7NextTop
>~~~Meetings:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[Note -- CFPs lately are getting too long! please send short
versions with all the wonderful details at your the conference website! GPS]
From: mpsingh@eos.ncsu.edu
Subject: 2nd CFP: Agent Theories, Architectures, and Languages, 1997 (4th Intl Wshop)
Date: Mon, 17 Feb 1997 18:20:54 -0500 (EST)
Reply-To: mpsingh@eos.ncsu.edu
SECOND CALL FOR PAPERS
The Fourth International Workshop on
Agent Theories, Architectures, and Languages (ATAL)
Intelligent agents are one of the most important developments in computer
science in the 1990s. Agents are of interest in many important application
areas, ranging from human-computer interaction to industrial process
control. The ATAL workshop series aims to bring together researchers
interested in the agent-level, micro aspects of agent technology.
Specifically, ATAL-97 will address issues such as theories of rational
agency, software architectures for intelligent agents, methodologies and
programming languages for realising agents, and software tools for applying
and evaluating agent systems. Papers that consider macro-level, societal
issues of agent-based systems are welcome only if they explicitly relate to
the workshop themes. ATAL-97 will be held over the three days immediately
preceding the AAAI-97 conference, also being held in Providence. The ATAL-97
proceedings will be formally published as volume four of the Intelligent
Agents series from Springer-Verlag.
TIMETABLE
Submissions due April 18, 1997
Notifications sent May 23, 1997
Prefinal versions due July 1, 1997
Workshop July 24-26, 1997
[edited for brevity -- full details at URL above. GPS] Previous8NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Mon, 17 Feb 1997 12:15:07 -0800 (PST)
From: H Michael Chung (hmchung@csulb.edu)
Call for Papers
Association of Information Systems 1997 Americas Conference
Indianapolis, Indiana, August 15-17, 1997
Mini-track on
'Tools and Applications of
Data Mining, Induction, and Knowledge Discovery:
In Search of a Mighty Tool'
Minitrack Chair: H. Michael Chung, CSULB
Description
This minitrack covers broader issues related to data mining, induction, and
knowledge discovery in the areas of business and management applications.
Tools based on regression analysis, information theoretic methods, genetic
algorithms, and neural networks have been applied to discover patterns of
financial fraud, to capture customer profiles for marketing, to predict
fluctuations in stock prices, to control product quality, and to diagnose
telecommunication network problems, among others . Expert decisions,
environmental/normative datasets, and Internet database are considered for
discovering information and knowledge.
There are many issues that should be addressed in order to reap quality
knowledge by applying sophisticated algorithms that would satisfy user
needs. Some of the relevant topics include
- Applications of Inductive Learning, Data Mining, and Knowledge
Discovery
- Data Warehousing
- Statistical Inference of Data Mining
- Knowledge Acquisition
- WWW Database and Agents
- Evaluation of Tools
- Economics of Decisions
- Data Visualization
- Learning Systems
Each submission must be FORWARDED ELECTRONICALLY AS A WORD PROCESSING FILE
(MS WORD OR WORDPERFECT FORMAT) ATTACHED TO AN E-MAIL MESSAGE to the
mini-track chair, H. Michael Chung. If this is not possible, then authors
should contact the mini-track chair and arrange for a suitable workaround.
Each submission is limited to THREE-PAGES IN LENGTH (APPROXIMATELY 1,750
WORDS) INCLUDING ALL FIGURES, TABLES, APPENDICES, AND REFERENCES, and must
include the
following:
a) The name, e-mail address, mailing address, university/organizational
affiliation, and phone/fax numbers of the contact person for the submission
in the first few lines of the file,
b) The submission title and the author's(s') name(s), the author's(s')
e-mail address(es), mailing address(es), and author's(s')
organization/university affiliation(s),
c) An abstract of the submission,
d) The body of the submission, and
e) A list of references or a bibliography.
All conference submissions and the submission review processes will be
managed through e-mail. The receipt of submissions will be quickly
confirmed by the mini-track chair. Submissions should follow the style
guidelines of the MIS Quarterly. All camera-ready copy preparation details
will be provided to submitting authors by the mini-track chairs through
e-mail upon acceptance.
Please send any questions and all submissions to Data Mining mini-track to
H. Michael Chung
Department of Information Systems
College of Business Administration
California State University, Long Beach
Long Beach, CA 90840-8506
TEL (562) 985-7691
FAX (562) 985-5543
INTERNET hmchung@csulb.edu
Previous9NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Date: Mon, 17 Feb 1997 15:05:21 +0100 (MET)
From: Luc De Raedt (Luc.DeRaedt@cs.kuleuven.ac.be)
CALL FOR PARTICIPATION and PAPERS
IJCAI-97 Workshop on
FRONTIERS OF INDUCTIVE LOGIC PROGRAMMING
Monday 25 August 1997
GENERAL INFORMATION
The IJCAI-97 one day workshop on 'Frontiers of ILP' in Nagoya, Japan,
will take place on August 25, immediately prior to
the start of the main IJCAI conference.
TECHNICAL DESCRIPTION
Inductive logic programming (ILP) is a recent subfield of
artificial intelligence that studies the induction of first order formulae
from examples. The purpose of this workshop is twofold:
on the one hand, we wish to widen the scope of ILP
by investigating its relations to neighboring fields,
and on the other hand, we wish to make ILP more accessible
for researchers from neighboring fields.
The workshop therefore solicits papers
that lie at the frontiers of ILP with neighboring fields.
A non-exclusive list of interesting topics for the workshop includes :
* ILP and Software Engineering:
what has ILP to offer to Software Engineering ?,
and in what way can Software Engineering help to design ILP systems
and applications ?
* ILP for Knowledge Discovery in Databases : ILP aims
at learning complex rules involving multiple relations from small
databases, whereas KDD typically induces simple rules about a
single relation from a large database. Furthermore, ILP allows to
exploit background knowledge in a variety of ways. Can KDD and ILP be
succesfully combined ?
* ILP and Computational or Algorithmic Learning Theory :
though many results have been obtained concerning the learnability
of inductive logic programming, most of the results are negative
and most of the positive results are reducible to propositional learning
methods. Is there a mismatch of COLT with ILP ? and if so,
what can be done about it ?
* ILP versus propositional learning methods :
Since the very start of ILP, researchers and practioners of
machine learning have wondered about the relation between
ILP and propositional learning methods. Theoretical and experimental
questions that arise include:
when to use ILP and when to use propositional learning methods ?
under what circumstances can ILP be reduced to propositional learning ?
what is the price to pay for using first order logic in
terms of efficiency ?
* ILP and Knowledge Representation : ILP has traditionally employed
computational logic to represent hypotheses and observations.
Alternative well-founded knowledge representation formalisms have received
little attention (with the exception of CLASSIC).
What can ILP learn from Knowledge Representation ?
and in what well-founded Knowledge Representation formalisms
is induction feasible ?
* ILP in multistrategy learning : Multistrategy learning
combines multiple learning strategies. What role can ILP
play for multistrategy learning ?
* ILP and Probabilistic reasoning: in contrast to
propositional learning methods, ILP has not used
probabilistic representations. How can ILP incorporate
such representations ? and how can it interact with
methods such as Bayes nets or Hidden Markov Models ?
* ILP for Intelligent Information Retrieval:
The rapid development of
the World Wide Web has spawned significant interest in intelligent
information retrieval. In particular, the need for algorithms for
reliably classifying textual documents into given categories (like
interesting/uninteresting) be useful for a wide variety of tasks.
Currently, most learning algorithms are not able to make use of
structural information like word order, succesive words, structure of
the text, etc. Can ILP algorithms offer advantages over conventional
information retrieval or machine learning algorithms for this sort of
tasks?
* Applications of ILP in subfields of AI : ILP has been applied
to other subfields of AI, including natural language processing,
intelligent agents and planning.
Further applications of ILP within AI are solicited.
Both position papers about the relation of ILP to other fields, as well
as research papers that make specific techical contributions
are solicited. However, to stimulate discussion, it is expected
that each technical paper also clarifies the position
of ILP with regard to the neighboring field(s) it addresses.
Except for the presentation of position and technical papers,
the workshop will also feature a panel discussion
on the frontiers of ILP and possibly an invited talk.
ORGANISERS
Luc De Raedt (chair and primary contact)
Saso Dzeroski
Koichi Furukawa
Fumio Mizoguchi
Stephen Muggleton
PROGRAMME COMMITTEE
Francesco Bergadano (Italy)
Luc De Raedt (co-chair, Belgium)
Saso Dzeroski (Slovenia)
Johannes Furnkranz (Austria)
Koichi Furukawa (Japan)
David Page (U.K.)
Fumio Mizoguchi (Japan)
Ray Mooney (U.S.A.)
Stephen Muggleton (co-chair, U.K.)
CALL FOR PARTICIPATION
Participation is open to all members of the AI Community.
However, to encourage interaction and a broad exchange of ideas
the number of participants will be strictly limited
(preferably under 30 and certainly under 40).
Participants will be selected on the basis of submissions.
Three types of submissions will be considered :
1) technical contributions (ideally, a 3 to 5 page extended abstract,
in the IJCAI Proceedings Format, 3000-4000 words),
2) position papers (ideally, a 1 to 3 page abstract
in the IJCAI Proceedings Format, 1000 - 3000 words)
3) a statement of interest (ideally, a one page motivation of why you
would like to participate, 300- 500 words)
Only submissions of type 1) and 2) will be considered
for presentation at the workshop and inclusion in the workshop notes.
Submissions should be received no later than April 1, 1997,
and must include first author's complete contact information,
including address, email, phone, and fax number. Though 1 April
is the hard deadline, the authors are encouraged to submit
their material by 24 March, in order to facilitate the reviewing process.
Double submissions with the ILP-97 Workshop (which is to take
place in Prague, September 1997) are allowed.
SUBMISSIONS
Submit papers by email (postscript) and surface mail (2 copies) to
Luc De Raedt
Dept. of Computer Science
Katholieke Universiteit Leuven
Celestijnenlaan 200A
B-3001 Heverlee
Belgium
Email : Luc.DeRaedt@cs.kuleuven.ac.be
IMPORTANT DATES
- Paper submission : 1 April
- Notification to Authors : 21 April
- Camera ready copy : the submissions themselve
will serve as camera ready copy
(submissions in the IJCAI Proceedings Style are strongly preferred,
see http://www.ijcai.org/ijcai-97/
for details)
PUBLICATION
The accepted submissions will be included in the workshop notes
to be distributed at the workshop.
Post-conference publication of a selection of the workshop papers
will be considered and discussed at the workshop.
COSTS
To cover costs, a fee of $US 50 will be charged,
in addition to the normal IJCAI-97 conference registration fee.
Attendees of IJCAI workshops will be required to register
for the main IJCAI conference.
Previous10NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: 'MANAGO' (manago@ibpc.fr)
Subject: 2 days course on Data Mining & CBR in San Francisco for University of Berkeley Extension
Date: Tue, 18 Feb 1997 17:23:09 +0100
Continuing Education in Engineering
University of California Berkeley Extension
Intensive short course at the San Francisco Airport
Course Organizer
Michel Manago, Acknosoft
Course Lecturers
Dr Usama Fayyad, Senior Researcher, Microsoft Research
Dr Michel Manago, President, Acknosoft international
Dr Evangelos Simoudis, Vice President, Data Mining and Decision Support
Solutions at IBM
Data Mining and Case-Based Reasoning (CBR): Principles and Applications
An intensive two-day course
Monday-Tuesday, March 24-25, 1997
San Francisco Airport
Course Description
The objective of this course is to present technologies for making better
use of data for decision-making purposes. Data mining techniques are used
to extract decision knowledge: for instance, in the form of a decision tree
or decision rules from a database. Case-based reasoning is the name given
to problem-solving methods that make direct use of past experiences (cases)
rather than a corpus of general knowledge. Data mining (DM) and case-based
reasoning (CBR) technologies can be used to:
* Explore and analyze databases and generate hypotheses about the data;
* Anticipate future events (decision support);
* Solve a new problem, whose solution is unknown, by retrieving and
adapting similar problems that have been previously solved.
According to the meta-group, the market for data mining is estimated at
$800 million by the year 2000. It is considered to be one of the three key
technologies that will have the biggest impact on information technologies
in the third millennium.
The course addresses both practical and theoretical issues. We will compare
and contrast the technologies, present the architecture of CBR and DM
systems, describe some algorithms, and more. We also will show how: cases
are indexed for efficient retrieval; the similarity between new and past
cases is assessed; cases can be represented; to use domain knowledge in
addition to data to characterize applications domains and reveal the
underlying methodology for building an application. We will identify the
market and present real applications in various domains such as technical
maintenance (diagnosis of Boeing 737 aircraft engines), customer support
(help desk for troubleshooting SEPRO robots in the plastic industry),
configuration (layouts of composite parts of an autoclave at Lockheed),
financial decision support, retail, and fraud detection.
Who Should Attend
This course is intended for:
* Business analysts who want to have an in-depth overview of data mining
technology and learn what it can really do and cannot do
* IT managers and technical staff who are in charge of engineering business
information systems and who want to learn how to implement data mining
solutions
* End-users who need to make better use of their data for decision making
* Customer service managers, maintenance managers, manufacturing managers,
financial decision makers who want to learn how to solve problems more
efficiently and at reduced costs
Anyone with a specific application in mind can benefit from the course,
which provides an overview of the technologies as well as of the
applications. Non-technical people will benefit from the basics of the
course, such as general principles and overview of applications
(quantification of business benefits, for example). There are no
prerequisites; this tutorial describes basic notions and illustrates these
with meaningful examples from a variety of applications in technical
maintenance, customer support, manufacturing, banking, and the consumer
market. Computer skills are not required.
Schedule
Monday-Tuesday, March 24-25, 1997
Registration: 8:00 am Monday
Lectures: 8:30 am-4:30 pm daily
Lunches: noon-1:00 pm daily
Location
Embassy Suites Hotel, San Francisco Airport, 150 Anza Blvd., Burlingame,
California.
Fee
The fee is $895 (EDP 326611). This includes:
* 2 days of instruction (1.4 ceu)
* Comprehensive course notes
* Daily lunches and refreshments
Topic Outline
Day One
From Data to Decisions
This brief introduction will provide to the attendees a common ground that
will enable them to understand and participate in the rest of the tutorial.
We will define knowledge discovery (KDD) in databases and case-based
reasoning (CBR)
Introduction to Knowledge Discovery in Databases
In this section we will:
Provide a general architecture for a generic KDD system that will enable
the subsequent discussion of the fundamental KDD issues, presentation of
the various KDD techniques, and description of various existing KDD
systems.
Present the basic knowledge discovery process, from the initial stages of
selecting data and cleaning of the selected data, to the identification of
important attributes and the final stages of integrating the extracted
knowledge into a decision support system.
Briefly discuss the various types of data mining techniques that are
commonly used for KDD. A brief introduction of CBR will be made.
Outline the core research issues in the field of KDD, as well as present
how these issues relate to fundamental AI issues such as representation and
search.
Preparing Data for Mining
The quality of the knowledge extracted by a KDD system from a data set is
related to the quality of the provided data. In this part of the tutorial
we will:
Examine various data problems, e.g., noisy data, incomplete data,
low-information content data, etc.
Discuss how each such problem affects the KDD operation.
Present techniques for solving certain of these problems, e.g., data
cleaning techniques. The large size of the databases that must be analyzed
necessitates the use of sampling techniques and the application of
dimensionality reduction techniques on a data set before a data mining
method is applied to it. We will present commonly used sampling methods and
discuss how they can be implemented. We will also discuss commonly used
dimensionality reduction techniques from statistics, e.g., principal
component analysis, and the use of domain knowledge for identifying
important attributes of a data set. Due to the particular prevalence and
importance of time-series data in a variety of application domains, we will
discuss techniques for preprocessing such data before it is presented to a
KDD system.
Data Mining and Technique Selection
We will present data mining techniques from five basic areas: (1)
artificial intelligence, (2) neural networks, (3) statistics, (4)
multidimensional and deductive databases, and (5) data visualization.
With each type of technique we will present its pros and cons with respect
to the generic KDD model defined in the tutorial's first part.
Databases and Visualization Techniques
Multidimensional and deductive databases merge knowledge-based techniques
with database technology. Recently such databases have been successfully
coupled with relational and legacy database management systems, providing
analysts with unique ways to express and automatically test hypotheses on
very large data sets. In addition, research on very large databases has
resulted in a variety of KDD techniques, such as association discovery and
sequence discovery. These techniques are based on simple database
operations, such as aggregation, and are applicable to specific types of
data, such as those commonly collected by large retail chains. We will
provide an introduction to multidimensional and deductive databases,
discuss data warehousing concepts, present how these techniques can be
applied on KDD tasks, and review the current research on databases.
Visualization has traditionally been used for the presentation of results
obtained by other methods, e.g., statistical analysis. We will discuss how
interactive visualization techniques can be used for knowledge discovery
operations. We will begin with simple techniques (scatter plots and line
plots, for example) and proceed with modern 3-D visualization techniques.
Some Examples of KDD Applications
We will first develop a set of criteria for comparing KDD systems. We will
then review in depth two such systems developed by the authors and
considered by the research community as representing the state-of-the-art:
IBM's customer segmentation data mining system and JPL's SKICAT system. In
addition to presenting the architecture of each system and discussing the
KDD methods it integrates, we will present a detailed account of how the
systems have been applied on financial, retail, manufacturing, astronomy,
and large image databases in planetary sciences.
Demonstration of a Data Mining System and Applications
Summary of the Day and Discussion
Summary, recap, overview of the basic unifying themes, and pointers to
available literature on KDD and future work.
Day Two
Overview of Case-Based Reasoning (CBR) Technology
In this introduction, we will present an overview of CBR, detail the CBR
cycle, and explain the main characteristics of CBR technology.
Applications of CBR in Technical Domains
We will present several CBR applications in technical domains. These deal
with maintenance, customer support, manufacturing, design, rapid evaluation
of production costs, and sale-support.
Troubleshooting CF56-3 engines for the Boeing 737. Time spent by airline
maintenance operators to solve engine failures and related costs (flight
delays or cancellations) are a major concern. The use of an intelligent
diagnostic software contributes to improving customer support and reduces
the cost of ownership by improving troubleshooting accuracy and reducing
airplane downtime. We will examine this application from the engine
manufacturer perspective (CFM international/Snecma) as well as from the
client's perspective (British Airways). Integration of the CBR
troubleshooting with electonic technical documentation. Demonstration.
A help desk for troubleshooting SEPRO robots in the plastic industry. Case
study from a small size company (160 employees) that has adopted CBR for it
customer support services. Demonstration.
Improving feedback from experience in manufacturing. We will present the
ongoing Noemie data warehousing and data mining project. Noemie aims at
increasing the quality and reliability of equipments for the oil industry.
Case study from the manufacturer perspective (Schlumberger) as well as from
the end-user's perspective (Nork Hydro).
CBR: How It Works
Based on the review of applications that will have been presented during
the morning, we will go into the details of the algorithms and present how
they have been used. In particular, we will describe mechanisms for:
retrieving cases; assessing the similarity; and indexing cases. We will
describe the link between induction, a form of KDD, and CBR. We also will
present some sample algorithms.
Comparing CBR with Other Technologies
During this part of the tutorial, we will compare CBR and other
technologies for decision making. In particular, we will look at rule-based
expert systems, classical statistics, neural networks, and standard
database queries. We will review a case study done at a banking institution
for comparing credit scoring, CBR, and rule-based expert systems.
Case-Based Reasoning in Practice
During this final presentation, we will detail the basic steps and a
methodology for building a CBR system. We will describe how to model cases,
state how cases can be acquired from scratch or from existing databases,
review potential sources for the cases, and explain how to choose an
algorithm. We will also investigate organizational issues for assuring case
quality and explain how human factors have to be taken into consideration
when delivering a CBR application.
Summary of the Tutorial and Discussion
Lecturers
Usama Fayyad, Ph.D., is a Senior Researcher at Microsoft Research. He is
also a Distinguished Visiting Scientist at the Jet Propulsion Laboratory
(JPL), California Institute of Technology, and an adjunct professor of
computer science at University of Southern California. Prior to joining
Microsoft Research, he headed the Machine Learning Systems Group at JPL and
was Principal Investigator of the Science Data Analysis and Visualization
Task and other tasks involving machine learning applications. He received
his Ph.D. in computer science and engineering from the University of
Michigan, Ann Arbor. He is a recipient of the NASA Exceptional Achievement
Medal (1994) and the 1993 Lew Allen Award for Excellence at JPL. He has
co-chaired Knowledge Discovery in Database conferences KDD-94 and KDD-95,
and is general chair of KDD-96. He is a co-editor of Advances in Knowledge
Discovery and Data Mining (AAAI/MIT Press 1996), and Editor-in-Chief of a
new journal on this topic (Kluwer).
Michel Manago, Ph.D., is the scientific and managing director of AcknoSoft.
Dr. Manago graduated from the University of Illinois at Urbana-Champaign
and obtained his Ph.D. at the University of Paris, writing his thesis on
'Integration of Symbolic and Numeric Techniques in Machine Learning.' He
has applied DM and CBR in technical domains such as diagnosis of Boeing 737
engines, customer support for marine diesel engines and robots, maintenance
of trains, reliability analysis of gas meters, experience feedback to
increase quality of production when manufacturing oil equipment, nuclear
safety, design of plastic parts in the manufacturing industry, and active
sale support over the Internet. He is author of the KATE line of products
for DM and CBR. He is editor of the book Advances in Case Based Reasoning
(Springer Verlag, 1995) and author of the report 'A Review of Industrial
Case-Based Reasoning. He received the Information Technologies European
Award in 1995 (the European 'Nobel prize' in computer technologies), among
other honors.
Evangelos Simoudis, Ph.D., is Vice President, Data Mining and Decision
Support Solutions at IBM, where he is responsible for the development and
deployment of data mining solutions to IBM's customers worldwide. Prior to
joining IBM, Dr. Simoudis was a Group Leader of the Data Comprehension
Group at the Lockheed AI Center where, since 1991, he led the development
and market introduction of the Recon data mining system and led research on
knowledge discovery in databases, machine learning, case-based reasoning
and their application to financial, retail, and fraud detection problems.
In 1994 Dr. Simoudis and his team were awarded Lockheed's Pursuit of
Excellence Award for their work on the Recon system. Dr. Simoudis is also
an adjunct assistant professor at the computer engineering department of
Santa Clara University. Dr. Simoudis holds a Ph.D. in computer science from
Brandeis University, an M.S. in computer science from the University of
Oregon, a B.S. in electrical engineering from the California Institute of
Technology, and a B.A. in physics from Grinnell College.
Enrollment Information
Enrollment may be made by companies or individuals. Enrollment is limited
and advance enrollment is required. Upon request, a place in the course
will be reserved for individuals who require time to obtain authorization.
To reserve a place, call (510) 642-4151, or fax (510) 642-6027.
How to enroll
By phone: You may enroll by phone if you use MasterCard, Visa, or American
Express; call (510) 642-4111.
By fax: If you use MasterCard, Visa, or American Express, fill out the form
on the back of this brochure and send it via fax number (510) 642-0374.
Please be sure to fax the entire form including the mailing label, if there
is one. Please provide all the information requested on the form.
By mail: Fill out and return the enrollment form provided.
By purchase order: Companies, agencies, and other organizations may pay
course fees by purchase order.
Enrollments must be accompanied by the full fee or by purchase order
authorization. You may pay by check or use MasterCard, Visa, or American
Express. Make checks payable to the UC Regents.
For efficient enrollment processing, we must have the Priority Code from
this publication, whether or not it is addressed to you. This five-digit
code (three numbers and two letters) appears on the mailing label above the
addressee's name. If there is no label on your copy, the code appears in a
box in the middle of the address surface.
Cancellation policy: Any cancellation is subject to a $30 processing fee.
Cancellations received less than five working days from the start of the
course are subject to a $100 cancellation fee. Substitutions may be made at
any time. If the course is not held for any reason, UC Berkeley Extension's
liability is limited to refund of the full course fee.
Confirming your enrollment: If you enroll by mail and have not received an
enrollment confirmation five days prior to the scheduled date of the
course, please call (510) 642-4151 to confirm that the course will convene
as scheduled.
Housing
A group of rooms will be set aside at the Embassy Suites Hotel, San
Francisco Airport, 150 Anza Blvd., Burlingame, California, and reservation
information will be sent to enrollees. Participants may reserve rooms in
advance with Embassy Suites, phone (415) 342-4600 or fax (415) 342-8109.
Special rates will be available; participants in these courses should so
identify themselves when requesting room reservations. Reservations must
be made no later than one month before the date of your course. After this
date room reservations will be accepted only on a rate and space
availability basis.
Airport transportation and parking
Courtesy shuttle service is provided between the hotel and the airport.
There is ample free parking available at the hotel.
Continuing education units (ceu)
These units are a nationally recognized means of recording noncredit study
and are accepted by many employers and relicensure agencies as evidence of
a serious commitment to career advancement and the maintenance of
professional competence. One ceu is awarded for each 10 hours of
attendance. If you want us to keep a record of your ceu study you must fill
out and return a form that will be distributed in class.
Program Coordinator
Linda Reid, Continuing Education in Engineering, University Extension,
University of California, Berkeley
Program Representative
Natalie Dennis, Continuing Education in Engineering, University Extension,
University of California, Berkeley
General Information
Housing
A group of rooms will be set aside at the Embassy Suites Hotel, San
Francisco Airport, 150 Anza Blvd., Burlingame, California, and reservation
information will be sent to enrollees. Participants may reserve rooms in
advance with Embassy Suites, phone (415) 342-4600 or fax (415) 342-8109.
Special rates will be available; participants in these courses should so
identify themselves when requesting room reservations. Reservations must
be made no later than one month before the date of your course. After this
date room reservations will be accepted only on a rate and space
availability basis.
Airport transportation and parking
Courtesy shuttle service is provided between the hotel and the airport.
There is ample free parking available at the hotel.
Continuing education units (ceu)
These units are a nationally recognized means of recording noncredit study
and are accepted by many employers and relicensure agencies as evidence of
a serious commitment to career advancement and the maintenance of
professional competence. One ceu is awarded for each 10 hours of
attendance. If you want us to keep a record of your ceu study you must fill
out and return a form that will be distributed in class.
Program Coordinator
Linda Reid, Continuing Education in Engineering, University Extension,
University of California, Berkeley
Program Representative
Natalie Dennis, Continuing Education in Engineering, University Extension,
University of California, Berkeley
If you have questions
Call (510) 642-4151, e-mail course@garnet.berkeley.edu,
fax (510) 642-6027,
or write to Continuing Education in Engineering, University Extension, UC
Berkeley,
1995 University Ave., Berkeley, CA 94720-7010
The University of California, in accordance with applicable federal and
state law and University policy, prohibits discrimination, including
harassment, on the basis of race, color, national origin, religion, sex,
disability, age, medical condition (cancer-related), ancestry, marital
status, citizenship, sexual orientation, or status as a Vietnam-era veteran
or special disabled veteran. This nondiscrimination policy covers
admission, access, and treatment in University programs and activities.
Inquiries may be directed as follows: sex discrimination and sexual
harassment: Carmen McKines, Title IX Compliance Officer, (510) 643-7895;
disability discrimination and access: Ward Newmeyer, A.D.A./504 Compliance
Officer, (510) 643-5116 (voice or TTY/TDD); age discrimination: Alan T.
Kolling, Age Discrimination Act Coordinator, (510) 642-6392. Other
inquiries may be directed to the Academic Compliance Office, 200 California
Hall, #1500, (510) 642-2795.
CONTRACT TRAINING
Enlist our experts at your location
At UC Berkeley Extension we're committed to working with you and your staff
to help achieve your objectives. Through the Berkeley Partnership for
Professional Development, we'll meet with you to analyze your staff's
training needs, then custom-design a program to satisfy your special
requirements. Or you can select from our many established courses.
Contract training offers:
_ Choice of format: from workshops and sequential classes to multiday
residential seminars
_ Highly qualified instructors
_ Convenient location: on-site at your company or at a facility of your
choice
_ Courses tailored to your needs
To discuss your training needs,
call Karl Johnson at (510) 642-4151 or fax (510) 642-6027
ENROLL BY FAX with MasterCard, Visa, American Express, or a company
purchase order: (510) 642-0374.
Or enroll by phone with MasterCard, Visa, or American Express: (510)
642-4111.
Please give us the Priority Code (see below) if you enroll by phone.
To enroll by mail, return this entire page. Please do not remove the
mailing label.
Mail to: Dept. B, UC Berkeley Extension, 1995 University Ave., Berkeley, CA
94720.
Name
last
first
middle
Position
Company name
BUSINESS ADDRESS
number
street
mail stop
city
state
zip
Daytime phone
Fax number
These numbers are requested so that you can be notified if there is a
change in the schedule or status of your course.
Priority Code 6 0 9 ___ ___
For efficient processing, we must have the Priority Code from this
publication, whether or not it is addressed to you. This 5-digit code (3
numbers and 2 letters) appears on the mailing label above the addressee's
name. If there is no label on your copy, the code appears in a box in the
middle of the address surface.
I enclose $ ___________to cover_______enrollments in:
_____ Data Mining and Case-Based Reasoning
$895
EDP 326611
To pay by check, make check payable to the UC Regents.
To use oMasterCard oVisa oAmerican Express check appropriate box and give:
account number
date card expires
authorizing signature
For companies/agencies:
_____ Purchase order enclosed
(For proper processing this form must accompany your purchase order.)
Michel Manago
AcknoSoft
58 rue du Dessous des Berges
75013 Paris - France
tel : (33 1) 44 24 88 00, fax : (33 1) 44 24 88 66
web : http://www.AcknoSoft.com
Previous11NextTop
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
From: 'MANAGO' (manago@ibpc.fr)
Subject: Tutorial on CBR & Data Mining in London + 2 days seminar
on applications of CBR & Data Mining
Date: Tue, 18 Feb 1997 17:32:40 +0100
The following events are taking place in London on 17-19 March 1997
For registration please see the website http://www.unicom.co.uk.
Principles & Applications of CBR & Data Mining
UNICOM Tutorial + Seminar Organized by Dr Michel Manago, Acknosoft
OBJECTIVES:
The objective of the tutorial is to present technologies for making better
use of data for decision making purposes. Induction is a data mining
technique that is used to extract decision Knowledge, for instance in the
form of a decision tree or decision rules, from a database. Case-Based
Reasoning is the name given to problem solving methods that make direct use
of past experiences (cases) rather than a corpus of general Knowledge. The
technologies can be used for:
1. Exploring and analysing databases and generate hypothesis about the
data
2. Anticipate future events (decision support)
3. Solve a new problem, whose solution is unknown, by retrieving and
adapting similar problems that have been previously solved.
During this course, we will describe the underlying techniques and
methodologies to improve the decision making process by making better use
of data. The course will address both theoretical and practical issues. We
will compare and contrast the technologies, present the architecture of a
CBR and a DM System, describe some algorithms etc. We will show how cases
are indexed for efficient retrieval, how the similarity between new and
past cases is assessed, how cases can be represented, how to use domain
knowledge in addition to data, characterise applications domains and reveal
the underlying methodology for building an application. We will identify
the market and delineate real applications in various domains.
A. From data to decisions
The brief Introduction will provide to the attendees a common ground that
will enable them to understand and participate in the rest of the tutorial.
We will define Data Mining (induction) and Case-Reasoning (CBR).
B. Introduction to induction
In this section we will:
1. Present how to generate decision tree by induction
2. Present the inductive process, from the initial stages of selecting data
to the identification of important attributes, and the final stages of
integrating the extracted knowledge into a decision support system.
C. Presentation of Based Reasoning (CBR) technology
In this introduction, we will present an overview of CBR, detail the CBR
cycle and explain the main characteristics of CBR technology..
D. CBR : how it works
We will go into the details of the algorithms and present how they have
been used. In particular, we will describe mechanisms for :
1. retrieving cases
2. assessing the similarity
3. Indexing cases. We will describe the link between induction, a form of
KDD, and CBR
Finally, we will present some sample algorithms.
E. Preparing Data for CBR and Data Mining
The quality of the knowledge extracted by a decision support system from a
data set, is related to the quality of the provided data. In this part of
the tutorial we will examine various data problems, e.g., noisy data,
incomplete data, low-information content data, etc.
F. Comparing induction and CBR with other technologies
During this part of the tutorial, we will compare KDD & CBR and other
technologies for decision making. In particular, we will look at rule based
expert systems, classical statistics, neural networks and standard database
queries. We will review a case study done at a Banking institution for
comparing credit scoring, CBR and rule base expert systems.
G. Applications of CBR and data mining
During this final presentation, we will detail the basic steps and a
methodology for building a CBR system. We will describe how to model cases,
stated how cases can be acquired from scratch or from existing databases,
review potential sources for the cases and explain how to choose an
algorithm. We will also investigate organisational issues for assuring case
quality and explain how human factors have to be taken into consideration
when delivering a CBR application. We will also try to characterise the
market for CBR and data mining.
H. Summary of the tutorial and discussion
PRESENTER:
Dr Michel Manago graduated from the University of Illinois in
Urbana-Champaign in 1983. He obtained his PhD in 1988 at University of
Paris on 'Integration of Symbolic and Numeric Techniques in Machine
Learning. Since 1991, Dr Manago has been the scientific and managing
director of AcknoSoft where he has been 'putting the technology to use'.
Michel Manago is the father of the KATE line of products for taking smart
decisions from data. He was chairman of the 2nd European workshop on CBR in
1994, editor of the book Advances in Case Based Reasoning (Springer Verlag,
1995) and author of the report 'A review of industrial Case Based
Reasoning. Dr Michel Manago received the Information Technologies European
Award in 1995 (the European 'Nobel prize' in computer technologies), the
1st prize for innovative software application at the XPS trade show in
Germany in 1995 and the 1996 Application of the Year award by the French
computer magazine 'Decision micros et rouseaux'.
CBR and Data Mining: Putting the Technology to Use
BACKGROUND
Companies have gathered vast amounts of data that is not well used. Some
corporate databases almost work in write-only mode! Well exploited, this
mass of data could be turned into strategic corporate knowledge :
- the marketing department wants to discover trends in buyer behaviour
- the after sales division must work more efficiently so that the company
keeps customers
- the financial department wants to assess risks in a better way
- quality management and control must be improved...
However, going from data to decisions is not an easy task.
Innovative computer technologies such as data mining and Case Based
Reasoning (CBR), will help you solve complex problems in domains where
experience plays a critical role in good decision making. And with only a
short delay develop a solution and a guaranteed payback.
(C) Copyright AcknoSoft, 1997
OBJECTIVES :
The goal of this seminar is to get a clear view about the state of the art
of applying data mining and CBR technologies to solving practical problems.
The emphasis of the seminar will be on presentations done by users of the
technology as opposed to technology providers. They will share their
experience and delineate the benefits as well as the difficulties of
putting the technologies into use. The themes that will be covered by the
speakers include
- What are CBR and data mining?
- Features of the software products they have used to build their
application
- Comparison of data mining and CBR with other technologies
- Methodologies for case acquisition and maintenance
- Ensuring case quality and monitoring it over time
- Organisational issues that needed to be solved in order to field the
application
- Human factors
- Overcoming technological risks
- Cost and benefits of using data mining and CBR in various domains
The goal of the seminar is to present a clear view about issues that are in
common when building CBR and data mining applications in different domains
(banking, insurance, customer support and help desk, manufacturing,
energy). We will focus on general topics such as how to assess the costs
and quantify the benefits of using the technology, how to model cases so
that they contain the right sort of knowledge for decision making purposes,
how to use the tools to build systems that analyse cases efficiently or how
to manage a CBR project from the customer's perspective.
Benefits of Attending
-Find out how the knowledge of your specialists available to everyone in
your organisation
-Learn how to solve problems more quickly without the burden of building
expert systems
-Capitalise your experience
-Elicit the user point of view
-Share experience with other CBR application developers
-Find out how to analyse and distill your data into usable knowledge
-Take smart decisions that are based on your experience
Programme
Day 1
Brief introduction by Michel Manago
Short presentation about Data Mining and CBR, introduction of the
objectives of the seminar.
Using data mining and CBR at Deloitte & Touche
Olivier Curet and Jonathan Killin Deloitte & Touche Consulting Group UK,=
.
This talk will deal with application of financial fraud detection as wel=
l
as
about how to transfer pricing and trade mission.
Managing CBR and other new technologies
Paul Laughlin, TSB General Insurance
This presentation discusses the introduction of new technologies in a
Financial Services company. The experience of successful management of AI
systems development is analysed. The lessons learnt are discussed and
guidelines set for similar projects.
Costs related to CBR development and how they changes when more experience
from CBR technology is gained.
Marko Vanska, Nokia Finland
Nokia has already successfully run 3 different projects and 8 projects are
ongoing. Marko Vanska presents Nokia's experience about wide usage of CBR
technology for troubleshooting telecommunication equipment.
Applying CBR at BT
Andy Jones, BT Global Sales & Services
In a world of explosive technological innovation, globalisation and
deregulation, cost effectively offering consistent expert service
world-wide has become a significant issue for BT. Several successful CBR
projects have been implemented and work is now underwayto deploy what will,
in time, become a CBR based corporate knowledge infrastructure. The key
challenge being faced is to move from the hand-crafted specialist systems
towards the integration of CBR into the mainstream business.
CBR for aircraft maintenance at British Airways
Rick Magaldi, British Airways
Building a corporate memory for decision making purposes is an important
issue at British Airways. Aircraft are equipment that run for a long period
of time and, with staff turnover, experience is lost quickly. In
particular, for Concorde aircraft, knowledge about maintenance must be
capitalised before skilled experts retire. Rick Magaldi presents the
current status of CBR applications development at British Airways.
Increasing case quality for maintaining aircraft engines of the Boeing 737
Richard Heider, SNECMA, France
Time spent by airline maintenance operators to solve engine failures and
the related costs (flight delays or cancellations) are a major concern to
SNECMA which manufacture engines for civilian aircrafts such as Boeing 737s
and Airbus A340s. The use of an intelligent diagnostic software contributes
to improving customer support and reduces the cost of ownership by
improving troubleshooting accuracy and reducing airplane downtime.The
presentation focusses on how SNECMA has improved the technical quality of
the cases for developing diagnostic software for all Boeing 737 engines.
A help desk for maintaining robots for manufacturing plastic parts.
Michel Jez, Sepro France and UK
Sepro is a small company that sells and support robots world-wide. These
are used for manufacturing plastic parts. Michel Jez will report how CBR
technology can be put into practical use for a company with 150 employees
and describe how to overcome human problems with technicians who are not
always experts in using computers.
Day 2
Sale support of electronic devices.
Sean Breen, IMS Ltd
Analogue devices produce electronic components that are sold worlwide. Each
component belongs to families of products. The specifications of each
product is slightly different and a parametric search, that relies on CBR,
is performed by the sale-support system to retrieve the products that best
match a customer's need.
Obtaining better feedback from experience for the Ariane IV space rocket.
Luc Bregault, Matra Cap Systemes, France
Better exploitation of incident reports, even minor ones, aims at improving
the safety and reliability of mission critical equipment used in the space
industry. CBR and data mining allow classification of incidents as well as
the retrieval of similar ones.
Improving the quality of mission critical devices for the oil industry with
data warehouses and data mining : the NOEMIE project.
Charles Durbec, Schlumberger
Manufacturing knowledge about equipment is disseminated across different
databases: manufacturing databases, engineering databases, operation
databases etc. When equipment breaks down, analysing the information that
is in different locations is a complex task. This session presents the
NOEMIE project, that aims at improving the quality of products used in the
oil industry.
CBR for nuclear power plant maintenance.
Jean Louis Bouchet, EDF, France (tbc)
Improving customer satisfaction with CBR
Peter Mortimer, Bull
In the face of competition and pressure on profit, Bull turned to customer
care and service. They chose a CBR tool and developed an expert system for
client support. This project management success story will be discussed.
Train Maintenance with CBR
Antonio Ruggieri, Ansaldo Transporti, Italy (tbc)
Will the paper break? Automatically learning to predict manufacturing
problems.
Rob Milne, IA Ltd
One of Europe's largest paper making plants with typical manufacturing
problems. What to do with lots of historical measurements? How to
automatically learn the characteristics of the available data? The user's
view of the requirement and solution.
A help desk for supporting CAD/CAM stations
Thomas Pantleon, Mercedes, Germany
Michel Manago
AcknoSoft
58 rue du Dessous des Berges
75013 Paris - France
tel : (33 1) 44 24 88 00, fax : (33 1) 44 24 88 66
web : http://www.AcknoSoft.com
Knowledge Discovery Nuggets 97:07