Introduction: Conceptual Models for Intelligent Information Systems

The rapid explosion of information systems from various domains and organizations has created problems for decision-makers not only when using these systems, but also when deriving useful information from these sources. As information systems grow in size, there is hidden information that cannot be accessed easily yet has a profound impact on operational effectiveness and organizational success. To remedy these situations, current research has been focused on several fronts: data mining to extract useful information from large data sources, methods to aid users in querying information systems such as English-like query formulation, cooperative query answering and associative query answering; techniques to integrate heterogeneous data sources and to resolve semantic conflicts; and proactive information systems that alert users for monitoring dynamic information changes. In this special issue, we shall present five papers that address some of these problems. Two papers are theoretical in nature, one paper describes the techniques in handling exceptions of the workflow process, and the last two papers describe the use of conceptual models in information systems. In active information systems, when a user receives an alert, it is useful to provide an explanation about the alert. The first paper, by Minock and Chu entitled, “Explanation Over Inference Hierarchies in Active Mediation Applications”, discusses the use of inference hierarchies in generating explanations for active databases. The authors show that the computational complexity of the algorithms for generating explanations is linear with respect to the total number of conditions in inference hierarchies. The paper by T.Y. Lin entitled, “Data Mining and Machine Oriented Modeling: A Granular Computing Approach” provides a machine-oriented data model for data mining to derive interesting properties from the stored data. Associative rules and decision rules can be computed from the model. Lin then generalizes data mining to a granular (cluster of relations) computing approach for practical applications. Workflow is often used to model a process of an enterprise. Such models may be used to estimate and predict performance of process. The paper by Luo, Sheth, Kochut, and Miller entitled, “Exception Handling in Workflow Systems” proposes a defeasible-workflow to support exception handling for workflow management. Case-based reasoning is used to enhance the exception handling capability. The paper entitled “Conceptual Models and Architectures for Advanced Information Systems” by Kerschberg and Weishar describes the use of conceptual modeling to integrate information from multiple data sources. An intelligent thesaurus is used to mediate the different information representations. A special export schema is proposed to accomplish cooperative query processing. The last paper by Bressan, Goh, Levina, Madnick, Shah, and Siegel entitled, “Context Knowledge Representation and Reasoning in the Context Interchange System” proposes a context mediator to provide semantic conflict resolution among multiple heterogeneous data sources. The system provides flat files, databases, and web services. This collection of papers provides merely a snapshot of the current state of the rich field of conceptual modeling for intelligent information systems. I hope you enjoy reading the papers. For more information on these subjects, interested readers should refer to various journals and proceedings on the topics of data mining, conceptual modeling, and knowledge and database systems published by such organizations as AAAI, ACM, IEEE, etc.