Decision Technology in Management Track
Track Chair
Daniel R. Dolk
Naval Postgraduate School
Monterey, California 939435103
Phone: (831) 656-2260, Fax: (831) 656-3573
dolker@redshift.com
Agent-based Simulation and System Dynamics
Description: At its most basic level, economics addresses the following problem - there are n agents interacting with each other in markets. We want to find out the equilibrium set of actions of each of the agents in each of the markets as well as the associated prices and quantities. As agents and markets increase in number, the computing power required to solve this problem increases exponentially, and inversely with the level of precision, that one wishes to achieve. It is clear that for an even moderately complex economic model there is very little hope of actually solving them. One solution to this conundrum is to have agent-based models. Here, the power and intelligence of real life markets is mimicked in the laboratory populated by agents that act in markets exactly as they are supposed to do in real life. Rather than trying to compute the solution to models, we actually let the market mechanism find one for us. The computing power required now only increases exponentially in the number of markets. This agent-based economy approach can be adapted to a very wide range of applications as has been shown in the work emerging from the Santa Fe Institute. A different approach to the same problem is provided by the modeling tradition of System Dynamics developed at MIT. SD modeling rests on the idea of feedback, or circular causality, inherent in nonlinear systems. The observer/SD modeler attempts to capture causal relationships between elements of a given system by describing its feedback structure with a multiplicity of positive and negative feedback loops. Carefully formulated models are able to simulate the behavior of a complex real system at an aggregate level. Identifying leverage points and applying structural changes in such SD models help pre-testing real system changes and implementing new policies. Agent-based and SD simulation models have a capacity to deliver overlapping and complementing insights when applied to the same problems.
This minitrack is a premier presentation forum for the latest ideas and results in the area of agent-based adaptive simulation and/or SD-based simulation. We seek research papers, case studies and practitioner reports relating to agent-based and/or SD-based simulation methods, environments, and methodologies. Contributions, in which both agent-based and SD modeling techniques are employed, are highly welcome.
Topics: Relevant topics for this minitrack include (but are not limited to):
- Cross-sectional Agent-Based and SD studies
- Business Applications and Case Studies
- Architectures and Environments
- Standards
- Experimental Studies
- Large Scale Simulations
- Web-Based Simulations
- Model Verification, Validation and Calibration
- Solution Comparisons
- Integration With Decision Models
- Constructive and Destructive Agents
- Artificial Organizations, Societies and Economies
Minitrack Chairs
Alok Chaturvedi
Krannert School of Management
Purdue University
West Lafayette, IN 47907 USA
Tel: (765) 494-9048
Fax: (765) 494-1526
alok@mgmt.purdue.edu |
Daniel R. Dolk
Naval Postgraduate School (Code IS/Dk)
Monterey, CA 93943
Tel: (831) 656-2260
Fax: (831) 656-3407
drdolk@nps.navy.mil |
Jan Dickieson
Office of Naval Research (Code 342)
800 No. Quincy Street
Arlington, VA 22217-5660
Tel:703 696-9660
Fax:703 696-9777
dickiej@onr.navy.mil |
Hans J (Jochen) Scholl
Rockefeller College of Public Affairs and Policy
University at Albany / SUNY
Albany, NY 12222
Tel: (518) 442-3937
Fax:(518) 442-3886
jscholl@ctg.albany.edu |
Data Mining, Knowledge Discovery, and Information Retrieval
Description: This minitrack covers the broad theory and application issues related to data mining, machine learning, knowledge acquisition, knowledge discovery, information retrieval, and inductive decision making. Both structured and unstructured data repositories including human expert decisions, environmental/normative data sets, large document collections, and web databases are considered. Theoretical and methodological exploration in the previous years motivates us to further investigate the various and richer data and knowledge representation schemes such as multimedia and/or geographic data applied to science as well as management domains.
Comprehensibility of data mining techniques, data visualization, task and model interaction, data quality assessment, interpretability, scalability, human factors and modeling, performance measurement and validation, acquisition of qualitative knowledge, feature selection, and the economics of decisions are some of the topics being sought. Examples of applications are point of sale data analysis for customized marketing, credit card as well as mobile telephone transaction data analysis for fraud detection, stock market prediction, book searching in electronic bookstores and auctions, and digital library catalogues among others. Models come from a variety of fields including statistical analysis, neural networks, fuzzy logic, intelligent agents, multidimensional analysis, data visualization, and decision trees, etc. Relevant software and economic issues will also be considered. Also, major data mining research programs from both academic and industry groups are encouraged to submit results that address real world problems.
Topics: Possible Topics may include, but are not limited to, the following areas:
- Measuring Model Performance
- Algorithms and Tools
- Task Characteristics
- Data Quality, Size, and Representation
- Multimedia and Temporal Data
- Qualitative Knowledge
- Knowledge Reuse
- Active Learning
- Human Factors and Environmental Issues
- Mining Methodology Issues
- Scalability and Preprocessing
- Webbased Data Mining
- Internet Interface with Information Retrieval
- Model Interpretation
- Sensitivity and Generalizability
- Economics and Cost of Errors
- Geographic Data
- Data Visualization
- Text Categorization and Summarization
- Cross Language Retrieval
- Speech and Broadcast Retrieval
- Application Case Studies
- Industry Projects
Minitrack Chairs
e-Services: Models and Methods for Design, Implementation and Delivery
Description: This minitrack begins with the recognition that software is no longer a product, but increasingly a service or product-service combination. e-Services refers to the emerging area of IS and IT services that are delivered electronically - typically through the Internet, wireless or land-based telecommunications networks. e-services must not only be designed, integrated, and delivered on highly compressed schedules, they must also be customized to specific needs of different organizational clients.
The development of e-services share characteristics of both information systems development (ISD) and IT services that combinatorially pose new challenges for effective, customer-centric delivery. Unlike traditional software, e-services are often subscription-based-an area in which research is clearly lacking and has been called for in the June 1998 special issue of Information Systems Research.
e-service development is a knowledge-intensive process wherein the service provider organization creates different, customized versions of the same service for diverse clientele through mixing and matching subsystems and components that might include platforms, software, hardware, telecommunications networks, and their interfaces.
e-services also entail a higher level of risk sharing between the provider and client, unlike conventional information systems where the risk (and sunk cost) is borne primarily by the client. e-services also run a higher risk of turnover and discontinuance behavior because the client organization is not locked into a specific system investment. For a complex set of technological components that must meld to form an e-service platform, we contend that it must be viewed as a complex system and as a line of products. Therefore, it is essential to support its design through intelligent decision support and knowledge management methods, and consider interplaying conceptual, engineering, and managerial issues in order to create effective and efficient platforms for e-service delivery.
The key objective of this minitrack is to invite original work and provide a forum for emerging research exploring various aspects behind the design, delivery, integration, and management of e-services. e-service development, management, delivery, and design is necessarily is cross-functional, multi-disciplinary activity and we contend that researchers in this area stand to gain through a collaborative exchange of multi-disciplinary views. This minitrack intends to provide such a forum.
Topics: To promote diversity in perspectives, we encourage research submissions research along three key dimensions of e-services, especially those focusing on application service providers (ASPs): (1) concepts, models, and design decision support mechanisms for e-services, (2) theory building or framework construction research through empirical, qualitative, or technical "proof of concept" system-design approaches, and (3) innovative applications and case studies highlighting challenges, issues and e-service solutions. Relevant topics for the minitrack include, but are not limited to the following:
1. e-services: Concepts, Modeling, Design, and Evaluation.
- Concepts and methods for designing and delivering e-services
- Theoretical frameworks and practical challenges, and research issues
- Decision support mechanisms, methods, and tools for e-service design
- Decision support for e-service delivery infrastructure design
- Tools and mechanisms for redevelopment cycle time compression
- Scalability, integration, and maintenance considerations mechanisms for e-service design in temporally and spatially distributed virtual teams
- Methods and concepts for one-to-many delivery and mass-customization
- Knowledge management across families of related e-services
- Design and process traceability for rapid design and delivery
- Management of e-services
- Component-based architectures and modularity in e-service design
- Relationships and comparison with information systems development (ISD) paradigms and theories
- Usage and measurement mechanisms for pricing and charging for e-services
- Outsourcing of IS services through the Internet
- Innovation management in the context of e-services
- Integrating enterprise resource planning (ERP) and e-service platforms and tools
- Theory building case studies of ASPs, ISPs, and e-commerce back office services
2. e-services: Applications.
- Innovative implementations of e-service delivery systems such as ASPs
- Single case studies and multi-case analyses of e-service implementation
- "Proof of concept" intelligent decision support tools for e-service design
- Practioner reports on e-service implementation trials and tribulations
Submissions focusing on e-services in the context of e-business applications for e-services are also welcome.
Minitrack Chairs
Balasubramaniam Ramesh
Department of CIS
J. Mack Robinson College of Business
Georgia State University
35 Broad Street
Atlanta, Georgia 30303
Tel: 404-651-3823
Fax: 404-651-3842
bramesh@gsu.edu
|
Amrit Tiwana
Department of CIS
J. Mack Robinson College of Business
Georgia State University
35 Broad Street
Atlanta, Georgia 30302
Tel: 404-651-3823
Fax: 404-651-3842
E-mail:atiwana@gsu.edu
|
Intelligent Systems and Soft Computing
Minitrack Chairs
Intelligent Systems in Traffic and Transportation
Description: The minitrack focuses on Intelligent Systems which are able to assist the design-phase (strategic planning) of traffic and transportation systems and/or the management-phase (tactical and operational planning) as well. The purpose of transportation logistics is to design, to organize and to manage transportation in order to meet customer service demands and cost and environmental requirements. Such logistics systems must comply with regulations on traffic, laws on labor and other types of constraints. In the field of transportation logistics we will focus on the analysis of urban, regional and intercity transportation networks for both passenger and freight transportation. Complex hybrid-type systems which include air-, road- and rail transportation are of particular interest.
Since the beginning of the last century an extraordinary development of transport demand is evident. This is a result of industrialization and the supply of new transport modes which at last made substantial changes of economy possible. The growing standard of living changed living and behavior patterns; faster and cheaper transport modes gave the impulse to see other regions, doing business with partners living farther away, who are able to offer goods cheaper than in the own region.
Therefore the increasing labor distribution in economy, the concentration of population in agglomerations with simultaneous migrations from rural regions, the demand for recreation for man being daily stressed in professional life as well as the improvement of rail, car, and air transport supply can be regarded as the principle causes for the enormous increase of demand.
To satisfy the increased demand for movement, to be economical with the use of public resources whilst encouraging economic growth, requires careful attention to monitoring trends and forecasting. Moreover, growth itself carries dangers. The expansion of transport facilities is not cost-free in terms of the environment and social welfare. A balance has therefore to be struck that satisfies a number of possibly conflicting goals. Such a balance requires that there are models forecasting the traffic demand and the changes in modal split and travel behavior.
Intelligent Systems which are designed to solve real world applications in traffic and transportation are built on the basis of an advanced software engineering concept including object-oriented software development and integration with non-standard databases and GIS. On the algorithmic side several so-called Intelligent Techniques coming from the AI, the OR and the CI, such as Tabu Search Metaheuristics, Evolutionary and Genetic Algorithms, Constraint Programming, but also high performance Optimization or Simulation techniques are used.
Topics: We seek research papers, case studies and practitioners reports relating to the design, the implementation and the use of Intelligent Systems built for particular problems in Traffic and Transportation. The papers need not to present fully developed complex systems. Conceptual papers, empirical papers and papers dealing with particular components of such a system, e.g. the Intelligent Techniques, are also welcomed.
Therefore, relevant topics for the minitrack include (but are not limited to)
- Modeling Intelligent Systems in Traffic and Transportation
- Models for the estimation of future volume of traffic likely to be affected by planned projects or management policy
- Models introducing changes in travel behavior
- Agent-based modeling and simulation of traffic related problems
- Transportation Network Design Problems including different modes of transportation and hubs
- Vehicle Routing and Crew Scheduling Problems (e.g. in air transportation)
- Supply Chain Management and Optimization
- Intelligent Techniques applied to combinatorial optimization problems in traffic and transportation logistics
- Tabu Search Metaheuristics
- Population-based methods (Genetic and Evolutionary Algorithms)
- Constraint programming
- Multi-Agent Approaches
- Conceptual papers on projects and reports on Intelligent Systems in practical use
Minitrack Chairs
Hans-Jürgen Sebastian
Aachen Institute of Technology
Operations Research
Templergraben 64
52056 Aachen, Germany
Tel.: +49-241-80 61 85
Fax: +49-241-8888-168
sebasti@or.rwth-aachen.de
|
Tore Grünert
Aachen Institute of Technology
Operations Research
Templergraben 64
52056 Aachen, Germany
Tel.: +49-241-80 61 87
Fax: +49-241-8888-168
tore@or.rwth-aachen.de
|
Modeling Knowledge-Intensive Processes: Concepts, Methods, and Applications
Description: This minitrack focuses on process knowledge, i.e., the processes involved in developing models, artifacts and decisions in complex organizational problem solving. We are interested in the concepts, rationale and methods that underlie the effective capture and dissemination of process knowledge. Our objective is to provide a continuing forum for emerging research on process knowledge, with particular emphasis on how diverse aspects of the problem can be integrated.
We encourage submissions on multiple aspects of the problem and seek to promote diversity in perspectives. Accordingly, the scope of the minitrack will encompass research on modeling concepts, methods, and applications. We also welcome submissions that focus on the use and efficacy of process knowledge in the design of products, systems or services.
Topics: Relevant topics for this minitrack include (but are not limited to) the following:
1. Modeling Process Knowledge: Concepts and Methods
- Concepts and methods for the capture and use of process knowledge in complex tasks such as design and delivery of products and services.
- Capturing and retaining "embedded" or tacit knowledge.
- Models and methods for linking process knowledge emerging from a particular design situation with the knowledgebase and memory of the larger organization.
- Factors influencing (impeding and facilitating) comprehensive capture and use of process knowledge.
- Frameworks for the evaluation of methods and tools.
- Tangible and intangible costs and benefits of managing process knowledge.
2. Modeling Process Knowledge: Applications
- Innovative implementations of "proof of concept" systems to support knowledge intensive processes.
- Systems for supporting virtual design teams.
- Case studies or experience reports on the capture and use of process knowledge in organizations.
- Empirical research on the use and efficacy of process knowledge in design and maintenance of products/systems.
- Systems to support rapid and evolutionary development.
- Traceability of process knowledge across system components and phases.
- Intelligent tools for management of process knowledge.
- Systems to support concurrent development.
- Process knowledge in formal software development.
Minitrack Chairs
Kishore Sengupta
Naval Postgraduate School
Monterey, CA 93943
Tel: 831 656-3212
Fax: 831 656-3407
kishore@nps.navy.mil |
Balasubramaniam Ramesh
Department of Computer Information Systems
College of Business
Georgia State University
35 Broad Street
Atlanta, Georgia 30302
Tel: 404-651-3823
Fax: 404-651-3842
bramesh@gsu.edu
|
Unified Modeling Language: A Critical Review and Suggested Future
Description: The Object Management Group (OMG) accepted the UML as an industry standard in 1998. UML is currently at release 1.3 and specifies a notation and modeling language for object oriented systems analysis and design. In addition, there is a formal specification language, Object Constraint Language (OCL). Adoption of UML in teaching and industry has been dramatic. There is clearly a need, which UML in some respects satisfies. Although there is a published meta-model, UML does not have a deep or consistent semantics behind the modeling language. The current specification also does not address the process or method dimension required to apply the UML in practice. A variety of methods, including the Unified Process from Ivar Jacobson and colleagues at Rational, are now emerging to address this aspect. None of these has yet been ratified as a standard.
Critical review of UML is beginning in earnest as industry tries to apply it in practical projects. Assessment of the variety of methods claiming support for the UML notation is urgently needed. Experience reports and suggestions for improvement can provide valuable input into the ongoing revision process under auspices of the OMG.
Topics of interest: To solicit high quality papers and innovative discussions in the areas of:
- Practical experience of UML usage
- Empirical research in the application of UML
- Critical examination of the UML and its associated meta-model
- Use of OCL and its integration with UML and support in tools
- Proposals for extensions of UML
- Critiques of, and proposals for, processes and lifecycles using the UML to express deliverables (e.g. Catalysis, Unified Process)
- Issues in the automation of UML and support in tools
- Metrics for UML models and the application of UML
- Comparison (particularly with respect to efficacy) between UML and competing notations/approaches (e.g., Open, ORM)
Minitrack Chairs
Hannu Kangassalo
University of Tampere
Kalevantie 4
33014 FINLAND
hk@cs.uta.fi |
Keng Siau
Department of Management
University of Nebraska-Lincoln
Lincoln NE 68588-0491
Tel: 402-472-3078
Fax: 402-472-5855
ksiau@unl.edu |