Machine Learning List: Vol. 15, No. 6 Thursday, April 24, 2003 Contents Calls for Papers and Other Meeting Announcements ECML/PKDD 2003 Discovery Challenge CfP Call for Papers: 4th EKDB 03 cfp: AMR 2003 - 1st Intl. Wrkshp on Adaptive Multimedia Retrieval CfP: KI-03 WS on Preference Learning CFP: ICML Workshop - Continuum from Labeled to Unlabeled Data ... CFP: Wrkshp on Probabilistic Graphical Models for Classification CFP: Operational Text Classification 03 : Wash., DC 27-Aug-03 Announcement of Workshops/Tutorials of ECML/PKDD-2003 CFP: MRDM Wshp @ SIGKDD-2003 Career Opportunities Position: Research Fellow in User Modeling Program Director for AI & Cognitive Science at NSF Misc. Other Announcements and Humor KDD CUP 2003 Announcement Announcement: Volume 1 of JMLG available. The Machine Learning List is moderated. Contributions should be relevant to the scientific study of machine learning. Please send submissions for distribution to: ml@isle.org. For requests to be added, removed, or to change your email address, send email to: ml-request@isle.org. In general, submissions should be no more than a few full-screens of text. For meeting announcements, highlight the conference or workshop web page and give a summary description of the goals of the event. Information such as the list of program committee members, talk schedules, and registration forms are unnecessary and should not be included. Job adds are usually no more than a few full-screens so they should fit naturally. ---------------------------------------------------------------------- From: "Petr Berka" Subject: ECML/PKDD 2003 Discovery Challenge CfP Date: Mon, 24 Mar 2003 13:20:34 +0100 ECML/PKDD 2003 Discovery Challenge Call for Contributions The Discovery Challenge will be held as a workshop at the 13th ECML + 6th PKDD conference, September 22-26, 2003, Cavtat-Dubrovnik, Croatia. Data from medical domain are available to prospective participants for download and analysis. The deadline for submissions is June, 30. For more info about the Challenge, visit http://lisp.vse.cz/challenge/ecmlpkdd2003 For more info about the ECML/PKDD2003, visit http://www.cs.kuleuven.ac.be/conference/ecmlpkdd/ ------------------------------ From: Subject: Call for Papers: 4th EKDB 03 Date: Mon, 24 Mar 2003 14:57:02 +0000 CALL FOR PAPERS 4th International Workshop on Extraction of Knowledge from Databases (EKDB'03) part of the 11th Portuguese Conference on Artificial Intelligence (EPIA'03) http://www.di.uevora.pt/epia03/ December 4-7, 2003 Beja, Portugal Aims and Scope The objective of the workshop is to discuss methods for non-trivial extraction of knowledge which is implicit in existing data and which can be described in a high-level representation so as to facilitate interpretation. Techniques from the machine learning, statistics and database fields are highly relevant for this task. Current real-world learning problems involve very large and complex data sets. Although a large number of techniques has been developed and applied, significant challenges remain, related with the design and analysis of methologies to handle this type of problems. The ability to incorporate new information and to react to concept drift, are challenging topics for different learning communities. One of the goals of this workshop is to promote an open discussion on these and related topics between different communities that are interested in these problems, namely the artificial intelligence, control, statistics and database communities. EKDB-03 follows the successful workshops EKDB-01, EKDB-99 and EKDB-97. IMPORTANT DATES May 18, 2003: Submission Deadline, July 20, 2003: Author Notification, September 13, 2003: Final versions due, December 4-7, 2003: Workshop and Conference ------------------------------ From: "Marcin Detyniecki" Subject: cfp: AMR 2003 - 1st Intl. Wrkshp on Adaptive Multimedia Retrieval Date: Thu, 27 Mar 2003 18:57:08 +0100 1st International Workshop on Adaptive Multimedia Retrieval - AMR 2003 - Part of KI 2003, 15-18 September 2003 University of Hamburg, Germany (http://www.cs.berkeley.edu/~anuernb/amr2003/) During the last years several approaches have been developed that tackle specific problems of the retrieval process, e.g. feature extraction methods for multimedia data, problem specific similarity measures and interactive user interfaces. These methods enable the design of efficient retrieval tools if the user is able to provide an appropriate query. However, user specific interests and search context are usually neglected when objects are retrieved. To improve today's retrieval tools and thus the overall satisfaction of a user, it is necessary to develop methods that are able to support the user in the search process, e.g. by providing additional information about the search results as well as the data collection itself and also by adapting the retrieval tool to the user's needs and interests. The goals of the workshop are to intensify the exchange of ideas between different research communities to enable the design of improved user adaptive retrieval tools. The workshop focuses especially on researchers that are working on feature extraction techniques for multimedia, computer linguistic approaches, (dynamic) data analysis methods, and visualization methods as well as user interface design. SUBMISSIONS: Submissions should be formatted according to Springer LNCS style (see http://www.springer.de/comp/lncs/authors.html). Papers should have about 10 pages but should not exceed 15 pages and should be submitted electronically in PDF or postscript. IMPORTANT DATES: May 24, 2003: Deadline for paper submission June 10, 2003: Notification of acceptance July 31, 2003: Deadline for final papers VENUE: The workshop will take place during the 26th German Conference on Artificial Intelligence (KI 2003) in Hamburg, Germany. Information about the venue, hotels, etc. are provided on the Web pages of the main conference KI 2003 (http://www.ki2003.de/). Further details can be found on the Web page of the workshop: http://www.cs.berkeley.edu/~anuernb/amr2003/ ------------------------------ From: johannes.fuernkranz@t-online.de (Johannes Fuernkranz) Subject: CfP: KI-03 WS on Preference Learning Date: Thu, 10 Apr 2003 09:56:32 +0200 Preference Learning: Models, Methods, Applications http://www.mathematik.uni-marburg.de/~eyke/Research/KI03WS.html A Workshop to be held as part of the conference KI-2003 , September 15-18, 2003, Hamburg WORKSHOP CONTENTS The focus of the workshop will be on machine learning methods for preference elicitation, i.e. on methods for inducing preferences from given observations. Like other types of complex learning tasks that have recently entered the stage in the field of machine learning, preference learning deviates strongly from the standard machine learning problems of classification and regression. It is particularly challenging because it involves the prediction of complex structures, such as weak or partial order relations, rather than single values. Moreover, training input will not, as it is usually the case, be offered in the form of complete examples but may comprise more general types of information, such as relative preferences or different kinds of indirect feedback. For example, learning problems might be posed by providing - or, in the style of an active learner, by asking for - preference relations between the training examples rather than a target value (as in supervised learning) or a utility degree (as in reinforcement learning). WORKSHOP GOALS The workshop pursues two main goals. Firstly, to discuss recent advances in preference elicitation through machine learning. Secondly, to stimulate new research avenues in this evolving field, by providing a discussion forum for both researchers in machine learning and potential users of preference elicitation techniques in all areas of Artificial Intelligence. Topics of interest include, but are not limited to * machine learning methods for preference elicitation, * extensions of the common frameworks for machine learning (supervised, unsupervised and reinforcement learning), * quantitative and qualitative approaches to preference modeling, * formal modeling of training examples and different forms of feedback, * applications of preference elicitation in various fields, e.g., in electronic commerce, personalization, or collaborative filtering. As the workshop is intended to support an exchange of ideas between different research areas interested in preference elicitation, it should appeal to both, researchers that work on preference elicitation techniques and tools (primarily people working in machine learning and related fields such as data mining, knowledge discovery, and statistics), as well as participants from other fields interested in preference elicitation (such as decision and game theory, autonomous/software/web agents, information retrieval, knowledge representation, negotiation, personalization, user modeling, or web intelligence). SCHEDULE Submission of extended abstracts: May 31, 2003 Acceptance notification: June 15, 2003 Final manuscripts: July 31, 2003 FURTHER INFORMATION Consult the Workshop home-page at http://www.mathematik.uni-marburg.de/~eyke/Research/KI03WS.html ------------------------------ From: rayid.ghani@accenture.com Subject: CFP: ICML Workshop - Continuum from Labeled to Unlabeled Data ... Date: Mon, 14 Apr 2003 13:28:25 -0500 CALL FOR PAPERS ICML 2003 Workshop (Co-located with KDD 2003) The Continuum from Labeled to Unlabeled Data in Machine Learning and Data Mining (Special emphasis on real-world applications and problems) August 21, 2003. Washington, DC. http://www.accenture.com/techlabs/icmlworkshop2003/ PAPERS DUE: May 1, 2003 WORKSHOP DESCRIPTION: There is a spectrum of ways to use data in machine learning and data mining. At the one end is completely unsupervised learning or clustering, and at the other end is supervised learning where the target output is known for every instance. This workshop aims to explore the space between these extremes, with particular attention to a variety of real-world applications. Papers addressing novel types of data, methods of diagnosing when unlabeled data will help and when it will hinder, and applying techniques across multiple application domains and multiple levels of supervision are particularly encouraged. Papers discussing the acquisition of labels from real-world experts in real-world data mining problems are also encouraged. Data mining practitioners working on real-world problems with large amounts of captured/stored data but a high cost labeling process are encouraged to submit problem descriptions and possible solutions. FOR MORE DETAILS, see http://www.accenture.com/techlabs/icmlworkshop2003 ------------------------------ From: Subject: CFP: Wrkshp on Probabilistic Graphical Models for Classification Date: Tue, 15 Apr 2003 12:33:29 +0200 (MET DST) WORKSHOP PROBABILISTIC GRAPHICAL MODELS FOR CLASSIFICATION during the 14th European Conference on Machine Learning (ECML) and the 7th European Conference on Principles and Practice of Knowledge Discovery in Databases (PKDD) September 23, 2003, Cavtat-Dubrovnik, Croatia Workshop web page: http://www.sc.ehu.es/ccwbayes/ecml-pkdd-03-workshop/call.htm SCHEDULE + Paper submission deadline - 13 June, 2003 + Notification to authors - 4 July, 2003 + Camera-ready papers - 11 July, 2003 + Workshop date - 23 September, 2003 SCOPE Probabilistic graphical model paradigm has become a popular tool for encoding, representing and handling uncertain knowledge in expert systems over the last decade. Applications of this paradigm include wide areas of the reality (medicine, agriculture, economy, bioinformatics...). Currently, interest is emerging within probabilistic graphical models to use them as a tool to induce supervised-unsupervised classification models. Taking the well-known naive-Bayes classifier as a basic, extensions and improvements of this simple but effective algorithm are being proposed from the field of probabilistic graphical models in the last ten years. The works of prestigious authors of the area of machine learning, have enhanced the role of probabilistic graphical models to solve classification tasks. Apart from the desired high accuracy of the model, these approaches offer the opportunity to graphically show the probabilistic relationships between domain attributes. Among these relevant approaches the selective Bayesian classifier, the Autoclass procedure, the tree-augmented network, or exact model averaging with naive Bayes can be cited. These works, coupled with the spectacular development of the Bayesian network paradigm, are opening a wide range of possibilities to adapt probabilistic graphical models to solve classification tasks. Contributed works in this workshop should ideally be in the area of new algorithmic, theoretical approaches and applications in the use of probabilistic graphical models to solve supervised and unsupervised classification tasks. ------------------------------ From: "Dave L" Subject: CFP: Operational Text Classification 03 : Wash., DC 27-Aug-03 Date: Thu, 17 Apr 2003 22:52:40 -0500 CALL FOR PARTICIPATION Third Workshop on Operational Text Classification (OTC-03) August 27, 2003 Washington, DC (co-located with KDD 2003 Conference) The OTC workshops feature talks and discussion by developers and users of text classification in a range of real-world settings. The 2003 workshop particularly encourages presentations on uses of text classification in text & data mining. However, ALL applications of text classification are of interest, including controlled vocabulary and web directory indexing, construction of specialized information feeds, information security, help desk automation, content filtering (e.g. spam, pornography), and alerting. Prospective speakers should submit an abstract (maximum 750 words) to otc2003submit@daviddlewis.com by June 8, 2003. Visit http://www.daviddlewis.com/events/otc2003 for more information, or write otc2003info@daviddlewis.com. ------------------------------ From: Luis Torgo Subject: Announcement of Workshops/Tutorials of ECML/PKDD-2003 Date: Tue, 22 Apr 2003 14:26:58 +0000 14th European Conf. on Machine Learning (ECML-03) and 7th European Conf. on Principles and Practice of Knowledge Discovery in Databases (PKDD-03) 22-26 September 2003, Cavtat-Dubrovnik, Croatia http://www.cs.kuleuven.ac.be/conference/ecmlpkdd/ WORKSHOPS W1: First European Web Mining Forum http://km.aifb.uni-karlsruhe.de/ws/ewmf03/ W2: Multimedia Discovery and Mining http://ai.ijs.si/Dunja/MultimediaMining03/ W3: Data Mining and Text Mining in Bioinformatics http://kd.cs.uni-magdeburg.de/ws03.html W4: Knowledge Discovery in Inductive Databases http://www.cinq-project.org/ecmlpkdd2003/ W5: Graph, Tree and Sequence Mining http://www.ar.sanken.osaka-u.ac.jp/MGTS-2003CFP.html W6: Probabilistic Graphical Models for Classification http://www.sc.ehu.es/ccwbayes/ecml-pkdd-03-workshop/call.htm W7: Parallel and Distributed Computing for Machine Learning http://www.fe.up.pt/~rcamacho/ECML03-W7.html TUTORIALS T1: KD Standards http://www.comp.rgu.ac.uk/staff/dw/kd_standards.html T2: Data Mining and Machine Learning in Time Series Databases http://www.cs.ucr.edu/~eamonn/ECML_PKDD_03.html T3: Exploratory Analysis of Spatial Data and Decision Making using Interactive Maps and Linked Dynamic Displays http://www.commongis.com/tutorial/tutorial-PKDD-2003.html T4: Music Data Mining http://www.soi.city.ac.uk/~geraint/conklin/ TUTORIAL/WORKSHOP COMBOS T/W1: Context-Free Grammar Learning http://ilk.uvt.nl/~mvzaanen/ECMLPKDD/index.html T/W2: Adaptive Text Extraction and Mining http://www.dcs.shef.ac.uk/~fabio/ATEM03/ CHALLENGE WORKSHOP ECML/PKDD2003 Discovery Challenge: A Collaborative Effort in Knowledge Discovery from Databases http://lisp.vse.cz/challenge/ecmlpkdd2003/chall2003.htm ------------------------------ From: Saso Dzeroski Subject: CFP: MRDM Wshp @ SIGKDD-2003 Date: Thu, 24 Apr 2003 12:52:07 +0200 CALL FOR PAPERS MRDM 2003 - 2nd Workshop on Multi-Relational Data Mining organised at the 9th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining August 24 - 27, 2003, Washington DC, USA PAPER SUBMISSIONS DUE: 6 June 2003 WORKSHOP WEBSITE: http://www-ai.ijs.si/SasoDzeroski/MRDM2003/ WORKSHOP DATE: 27 August 2003 Multi-Relational Data Mining (MRDM) is the multi-disciplinary field dealing with knowledge discovery from relational databases consisting of multiple tables. Mining data which consists of complex/structured objects also falls within the scope of this field, since the normalized representation of such objects in a relational database requires multiple tables. The field aims at integrating results from existing fields such as inductive logic programming, KDD, machine learning and relational databases; producing new techniques for mining multi-relational data; and practical applications of such tecniques. The aim of the workshop is to bring together researchers and practitioners of data mining interested in methods for finding patterns in expressive languages from complex/multi-relational/structured data and their applications. TOPICS OF INTEREST The topics of interest (listed in alphabetical order) include, but are not limited to, the following: - Applications of (multi-)relational data mining - Data mining problems that require (multi-)relational methods - Distance-based methods for structured/relational data - Inductive databases - Kernel methods for structured/relational data - Learning in probabilistic relational representations - Link analysis and discovery - Methods for (multi-)relational data mining - Mining structured data, such as amino-acid sequences, chemical compounds, HTML and XML documents, ... - Propositionalization methods for transforming (multi-)relational data mining problems to single-table data mining problems - Relational neural networks - Relational pattern languages We also encourage submissions which present early stages of research work, software, and applications. ------------------------------ From: "Geoff Webb" Subject: Position: Research Fellow in User Modeling Date: Tue, 25 Mar 2003 15:42:24 +1100 Research Fellow in Computer Science School of Computer Science & Software Engineering a.. Department/Faculty: School of Computer Science & Engineering b.. Location: Clayton campus c.. Closing Date: 09/04/03 Applications are invited from qualified people for an appointment of Research Fellow for eighteen months in the area of user-modeling of web-site users. The applicant should have a PhD in Computer Science or a related field. The applicant should be familiar with user-modeling, data-mining, or machine-learning techniques, and should have good programming skills. Programming in C++, C, or Java is an advantage. The Benefits: $54,864 - $65,152 p.a. Level B Location: Clayton campus Contact: Professor Geoff Webb, Tel. 9905 3296 or email webb@infotech.monash.edu.au for inquiries and information. Applications: Professor G Webb, School of Computer Science & Software Engineering, Monash University, Vic 3800 or email as above by 9/04/2003. Quote Ref No. A034226 and include curriculum vitae and the names (with phone and facsimile numbers) of three referees in your application. ------------------------------ From: "Pazzani, Michael J." Subject: Program Director for AI & Cognitive Science at NSF Date: Tue, 15 Apr 2003 10:54:02 -0400 The Information and Intelligent Systems Division of CISE at NSF is recruiting a program director for the Artificial Intelligence & Cognitive Science Program. The name of this program was recently changed from Knowledge and Cognitive Systems to more accurately reflect the research it supports. The program covers areas of AI & Cognitive Science including Planning, Knowledge Representation, Machine Learning, Automated Reasoning, Integrated Agents and models of cognitive processes. The program director may be hired as a NSF employee or as a "rotator" from a university or government position. One mechanism to hire a "rotator" is to make a grant to your home institution that pays your 12-month salary, with NSF providing additional funds for housing and travel back to your home institution. Applications are due May 15, 2003. See http://www.cise.nsf.gov/vacn/index.html for more details and the application process. The Artificial Intelligence & Cognitive Science Program is one of several related programs in the division of Information and Intelligent Systems, which also includes programs in Robotics and Computer Vision, Human Language and Communication, Human Computer Interaction, Information and Data Management, Digital Libraries, Digital Society and Technologies, Universal Access, and Data and Application Security. Qualified persons who are women, ethnic/racial minorities, and persons with disabilities are strongly encouraged to apply. The National Science Foundation is an Equal Opportunity Employer committed to employing a highly qualified staff that reflects the diversity of our nation. If you'd like more information or would like to recommend someone to me, please contact me at the address below. Michael J. Pazzani Division Director, Information and Intelligent Systems National Science Foundation 4201 Wilson Boulevard, Suite 1115, Arlington, VA 22230 Bus: 703-292-8930 Bus Fax: 703-292-9073 E-mail: mpazzani@nsf.gov http://www.cise.nsf.gov/iis ------------------------------ From: Osmar Zaiane Subject: KDD CUP 2003 Announcement Date: Mon, 7 Apr 2003 16:36:34 -0600 (MDT) KDD Cup 2003 (http://www.cs.cornell.edu/projects/kddcup/index.html) The Ninth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2003) Every year, in conjunction with the ACM SIGKDD conference, a knowledge discovery and data mining competition (KDD Cup) is held to challenge the research community in industry and academia. This year's competition focuses on problems motivated by network mining and the analysis of usage logs. Complex networks have emerged as a central theme in data mining applications, appearing in domains that range from communication networks and the Web, to biological interaction networks, to social networks and homeland security. At the same time, the difficulty in obtaining complete and accurate representations of large networks has been an obstacle to research in this area. This KDD Cup is based on a very large archive of research papers that provides an unusually comprehensive snapshot of a particular social network in action; in addition to the full text of research papers, it includes both explicit citation structure and (partial) data on the downloading of papers by users. It provides a framework for testing general network and usage mining techniques, which will be explored via four varied and interesting task. Each task is a separate competition with its own specific goals. To learn more about the KDD cup competition rules, the tasks, and the datasets, please visit the KDD cup web site managed by the KDD cup chairs Johannes Gehrke, Paul Ginsparg and Jon Kleinberg. For more information, please refer to the SIGKDD Conference web site http://www.acm.org/sigkdd/kdd2003/ or go directly to the KDD Cup 2003 web site http://www.cs.cornell.edu/projects/kddcup/index.html ------------------------------ From: "Cycle L. Bittmap, Ph.D." Subject: Announcement: Volume 1 of JMLG available. Date: Tue, 01 Apr 2003 05:30:03 -0800 To the AI and ML community: We are pleased to announce the first volume of a new online journal, the Journal of Machine Learning Gossip, (http://www.jmlg.org/papers.htm). The volume includes the following award-winning papers: Markov Indecision Processes: A Formal Model of Decision-Making Under Extreme Confusion by Harry Q. Bovik, Judy Q. Goldsmith, Andrew Q. Klapper, and Michael Q. Littman Data Set Selection by Doudou LaLoudouana and Mambobo Bonouliqui Tarare On the Origin and Destiny of Inductive Machine Learning by Terran Lane Visit our web site jmlg.org/papers.htm for access to these papers and for more information about the journal and its goals. We look forward to serving you in the coming years. Sincerely, Cycle L. Bittmap, PhD on behalf of the editors of the JMLG ------------------------------ End of ML-LIST Digest Vol 15, No. 6 ***********************************