Applications and Innovations in Intelligent Systems XII - Proceedings of AI-2004, the Twenty-fourth SGAI International Conference on Innhovative Techniques and Applications of Artificial Intelligence

APPLICATION PROGRAMME CHAIR'S INTRODUCTION

ACKNOWLEDGEMENTS

APPLICATIONS PROGRAMME COMMITTEE

CONTENTS

BEST APPLICATION PAPER

A Case-Based Technique for Tracking Concept Drift in Spam Filtering

1 Introduction

2 Spam Filtering and Machine Learning

3 The Problem of Concept Drift

3.1 Definitions and Types of Concept Drift

3.2 Approaches to Handling Concept Drift

4 A Case-Based Approach to Concept Drift

4.1 Feature Selection

4.2 Case Retrieval

4.3 Case-base Management

5 Evaluation

5.1 Experimental Setup

5.2 Evaluation Metrics

5.3 CBR vs. Naive Bayes

5.4 Level 1 Learning - Continuous Updating with new Instances

5.5 Level 2 Learning - Model Rebuild with Feature Reselection

6 Conclusions

References

SESSION 1 : SYNTHESIS AND PREDICTION

Matching and Predicting Crimes

1. Introduction

2. Data and Initial Analysis

3. Matching Crimes

3.1 Single Crimes with Similarity-Based Retrieval

3.2 Naive Bayes Classifier and Spatio-Temporal Features

3.3 Matching Series of Crimes with Logic

4. Predicting Crimes

4.1 Survival/Failure Time Analysis

4.2 Bayesian Network

5. Conclusions and Future Directions

Acknowledgements

References

Story Plot Generation based on CBR

1 Introduction

2 Related Work

2.1 Computational Plot Generation

2.2 Natural Language Generation

3 A CBR System for Composing Plot Plans

3.1 The System Knowledge

3.2 The CBR Module

4 The Natural Language Generation Module

4.1 Input Data for the NLG Module

4.2 Stages of the NLG Module

5 Conclusions

Acknowledgements

References

Studying Continuous Improvement from a Knowledge Perspective

1. Introduction

1.1 Bacl^round for the Study

1.2 Why Apply Knowledge Structure Mapping to Continuous Improvement?

2. The Knowledge Structure Mapping

2.1 Outputs from a KSM Investigation

2.2 The Value of a Knowledge Centred Approach

3* The CI Investigation

3.1 Defining the Study Focus

3.2 Forming the Team

3.3 The Study Approach

3.4 Analysing and Evaluating Results

3.5 Report Delivery

4. Project Results

4.1 The Results

4.2 Analysis of the Results

4.3 Observations and Recommendations

5. Initial Business Reaction

6. Conclusion

References

ReTAX+: A Cooperative Taxonomy Revision Tool

1. Introduction

2. Related Work

3. System Architecture

4. Definitions for Ontological Analysis

4.1 Taxonomy

4.2 Class & Attributes

5. Operations on Taxonomy

5.1 Maintenance of x and dri

5.2 Manipulation of Attributes

5.3 Manipulation of Classes

5.4 Refinement Strategies

6. Implementation

6.1 Planned Enhancements

7. Conclusion

SESSION 2: SCHEDULING AND SEARCH

A Non-Binary Constraint Ordering Approach to Scheduling Problems

1 Introduction

2 Definition and Algorithms

2.1 Variable Ordering

2.2 Value Ordering

3 Example of Non-binary Scheduling Problem

4 Constraint Ordering Heuristic (COH)

4.1 Specification of COH

5 Analysis of COH

6 Evaluation of COH

7 Conclusion and future work

References

A Heuristic Based System for Generation of Shifts with Breaks

1 INTRODUCTION

2 PROBLEM DESCRIPTION

3 APPLICATION DESCRIPTION

3.1 Generation of breaks

4 APPLICATION USE

4.1 Definition of temporal requirements

4.2 Constraints regarding shift types

4.3 Weights of criteria

4.4 Definition of break types

4.5 Generation of shifts

5 CONCLUSION

Acknowledgment

References

A Topological Model Based on Railway Capacity to Manage Periodic Train Scheduling

1 Introduction

2 Problem Topology

2.1 Railway Traffic Rules, topological and requirement constraints

2.2 General System Architecture

3 Railway Capacity

4 Topological Constraint Optimization Technique

4.1 Topological Technique

5 Evaluation

6 Conclusions

References

Collaborative Search; Deployment Experiences

1 Introduction

2 Repetition &: Regularity in Web Search

3 A Review of Collaborative Search

4 I-SPY

4.1 Creating a Community

4.2 Example Session

5 Live-User Evaluation

5.1 Methodology

5.2 Test Scores

5.3 Group Differences

5.4 Result Selections

5.5 Summary

6 Lessons Learned

6.1 Exploiting Query Similarity

6.2 Query Recommendation

6.3 Result Diversity

7 Conclusions

References

SESSION 3: DIAGNOSIS AND MONITORING

A Model-Based Approach to Robot Fault Diagnosis

1 Introduction

2 Problem Formulation and Solution

2.1 Robotic Fault Description

2.2 Robot Diagnosis System

3 Robotic Fault Reasoning

3.1 Robotic Bounds Generation

3.2 Interval Filter

3.3 Component-Based Robotic Reasoning

4 Case study

4.1 Simulation results

4.2 Fault report

5 Conclusion

Acknowledgements

Notation

References

Automating the Analysis and Management of Power System Data using Multi-agent Systems Technology

1. Introduction

2. Post-fault Analysis and Management of Power System Data related to Protection Operation

2.1 Monitoring Systems and Data sources

2.2 Post-fault Analysis using Existing Software Tools

3. Protection Engineering Diagnostic Agents

3.1 Agents and their Interactions

4. Industrial Implementation of PEDA

4.1 Issues arising during industrial implementation

4.2 Resolving issues related to industrial implementation

4.3 Agents and their interactions

5. Concluding Remarks

References

The Industrialisation of a Multi-Agent System for Power Transformer Condition Monitoring

1 Introduction

2 Transformer Condition Monitoring

3 COMMAS

3.1 Functional Design

3.2 Detailed Design

4 Aspects of Extensible Design

4.1 The Java Agent DEvelopment Framework (JADE)

4.2 Ontology and Content Language

4.3 Raw Data Formats

5 Running COMMAS

6 Conclusion

References

SESSION 4: CLASSIFICATION AND DESIGN

An Improved Genetic Programming Technique for the Classification of Raman Spectra

1 Introduction

2 Description of Task

3 Analysis Techniques

4 Experimental Results

5 Conclusions & Future Work

PROPOSE - Intelligent Advisory System for supporting Redesign

1. Introduction

2. Analysis-Based Optimisation in Design Process

3. Development of the Intelligent Redesign System

4. The Knowledge Base

5. The Shell of the System

6. Evaluation of the System

Conclusions

Acknowledgements

References

The Designers' Workbench: Using Ontologies and Constraints for Configuration

1. Backround

2. Aims of Designers' Workbench

3. Related work

3.1 Constraint-based approaches

3.2 Ontology-based approaches

3.3 Combining constraints and ontologies

4. An illustrative example

5. Functionality

6. Additional features of the Designers* Workbench

6.1 Graph-based display of configuration

6.2 Checking incomplete configurations

6.3 Constraint rationales

7. Preliminary evaluation

8. Future work

8.1 Propagation of values

8.2 Functional descriptions

8.3 Design rationale storage (argumentation)

8.4 Constraint input

8.5 Ontology change

9. Summary

10. Acknowledgements

A Visualisation Tool to Explain Case-Base Reasoning Solutions for Tablet Formulation

1 Introduction

2 Solutions Require Explanations

3 What Needs to be Explained

3.1 Knowledge Containers

3.2 The Proposed Solution

4 Textual FORMUCASE

5 FORMUCASEVIZ

5.1 Ordering the Attributes

6 User Evaluation of FormuCaseViz

7 Related Work

8 Conclusions and Future Work

Acknowledgments

References

SESSION 5: ANALYSIS AND EVALUATION

Formal Analysis of Empirical Traces in Incident Management

1. Introduction

2. The Hercules Disaster

3. The Dakota Disaster

4. Categorisation of Error Types

5. Modelling Approach

6. Formalisation of an Empirical Trace

7. Validation of a Trace

8. Discussion

References

Modelling Expertise for Structure Elucidation in Organic Chemistry Using Bayesian Networks

1 Introduction

2 Related Work

3 Bayesian Networks

4 Modelling

4.1 Spectral and Structural Subunits

4.2 Aromatic Carbons

4.3 Laws of the Domain

4.4 Assignment of Probability Values

5 Results and Outlook

References

Evaluation of a Mixed-Initiative Dialogue Multimodal Interface

1. Introduction

2. System Architecture

3. Operation Overview

4. Comparison between the Mixed-initiative and Directed-Dialogue Interfaces

5. Accuracy and Usability Test

5.1 Participants Selection

5.2 Experimental Setup

5 3 Statistics Methods for Test Results

5.4 Speech Recognition Accuracy Result

5.5 Usability Attributes Result

5.6 Results according to user age

5.7 Error Recovery Test Result

6. Conclusion and Discussion

Acknowledgements

References

AUTHOR INDEX