Showing papers by "Riichiro Mizoguchi published in 1999"

Communication Content Ontology for Learner Model Agent in Multi-agent Architecture

Abstract: In this paper we describe domain-independent concepts in message content in a multi-agent architecture of Intelligent Educational Systems (IES) Although it is well accepted that a common ontology holds the key to fluent communication between agents, most researchers believe that the common ontology is domain ontology This has been one of the causes of the fact that the work on agent communication has concentrated on general-purpose languages with communication performatives In IES research, however, we can find task-dependent but domainindependent content, which we could “ontologize” By domain, we here mean subject domains such as geometry, chemistry, and mathematics, etc In the multiagent architecture of IES, we examined the information exchanged between the learner model agent and other agents, and abstracted the domain-independent concepts We represent them explicitly and domain-independently, to build an ontology By so doing, we believe that we can move toward a standard for the communication between agents in educational systems, so that the modules/agents can be reused In this paper we discuss in detail the ontological issues in the message content, and give some examples to show the ontology, that we have designed

An Ontology-Aware Authoring Tool - Functional structure and guidance generation -

Abstract: This paper describes an ontology-aware authoring tool for intelligent training systems. The major advantages of the authoring tool are: (1) To provide human-friendly primitives in terms of which users can easily describe their own models of a task (descriptiveness, readability). (2) To simulate the abstract behavior of the model in terms of conceptual level primitives (conceptual level operationality). (3) To verify the consistency of models. In this paper, we show the total image of an ontology-aware authoring system, describe the functional structure and the guidance generation mechanism in detail with an emphasis on the roles which the ontology plays. Finally, we introduce the intelligent training system SmartTrainer's Authoring Tools (SmartTrainer/AT) and its realization as an example.

An Ontological Analysis of Fault Process and Category of Faults

Abstract: In an interactive system for problem solving, its transparency concerning capability and limitation is critical for ease of understanding to facilitate the utility of the system. The capability of a model-based system largely depends on the characteristics of the model used in it. Our aim here is explication of capabilities and limitations of models in model-based diagnostic systems, in particular limitation of the “depth” of the cause of faults identified by a diagnostic mechanism using a kind of model. We focus on the process in which faults are induced and establish an ontology of faults including several categories of faults. It provides us with a vocabulary for specifying the scope of a diagnostic activity performed by a reasoning mechanism using a kind of model. Firstly, the ontology enables us to specify the performances of the models in diagnostic systems, which makes the systems transparent to users who want to know what type of faults the system can diagnose when they describe a target model for a reasoning mechanism. Next, we develop an interactive system to enumerate the “deeper causes” of a malfunction. The ontology enables the human users to control the diagnostic system in a plausible-first way by relaxing diagnostic assumptions interactively

Meta-Functions of Artifacts

Abstract: This paper proposes a new category of functions of artifacts called meta-function. Although conventional research on functional representation defines function as some kinds of abstraction of changes in objects associated with components, a meta-function of a component represents a role played by a function of the component for another function of another component without mention of such changes specific to the target system. The meta-functions explain the types of collaboration and dependency among functions of components as part of the design rationale of the target artifacts. This paper defines nine types of the meta-functions such as ToEnable and ToDrive, which form part of an ontology of functional concepts. The utility of the ontology in explanation and redesign is also discussed.

An ontology more than a shared vocabulary

Abstract: This paper discusses ontology-awareness of authoring tools. The most important role of ontology is to lay the theoretical foundation for IES development process. It maintains continuity from author”s conceptual understanding of educational task to the computational semantics of IESs. It provides human friendly vocabulary for authors to describe the educational task. For the authoring tools, on the other hand, it specifies the computational semantics of vocabulary. The goals of our research on ontology are to exemplify the benefits of ontology more than vocabularies through the development of an ontology-aware authoring tool for intelligent educational systems.

Towards Redesign based on Ontologies of Functional Concepts and Redesign Strategies

Abstract: This research aims at developing a redesign problem solver which proposes suitable modifications of a given artifact for a new requirement or an inconvenience. We concentrate on the two issues, capturing design rationales of the given artifact and organizing general redesign strategies. For the former issue, we propose an ontology of functional concepts, which enables the redesign system to identify functional structures representing a part of design rationales automatically. It particularly includes a new category of functional concepts called meta-functions for representation of dependency among functions of components. For the latter issue, we describe an ontology of the redesign strategies. Some general strategies for proposing redesign solutions in terms of the concepts in the ontology are shown. It reveals conceptualization behind the redesign strategies and then helps us specialize the Soar architecture into a universal redesign engine.

Human media interface system for the next generation plant operation

Abstract: An overview is given of a project on "Development of a human interface for the next generation plant operation" running as a subproject of the MITI-funded Human Media project. The goals of the project include developing a next generation human centered interface for plant operation. We adopted an oil refinery plant as a task domain and have developed an interface system consisting of: an Interface Agent which is responsible for plant monitoring and for direct interaction with operators; a Virtual Plant Agent responsible for 3D virtual display of the plant; a Semantic Information Presentation Agent for presentation of semantically interpreted information based on a sophisticated model of the plant; an Ontology Server for standardizing of Agents' understanding of the plant; and a Distributed Collaboration Infrastructure for enabling agents to perform collaboration. Design philosophies and a project overview are presented.

An ontology-based human friendly message generation in a multiagent human media system for oil refinery plant operation

Abstract: The role of an ontology and ontology server in the interface for oil refinery plant operation is described. It plays a central role as the knowledge media which is one of the key technologies in the Human Media project. Plant ontology as well as operation task ontology has been developed. The main task of the ontology server in the current implementation is message generation with appropriate word selection which is done by a focused point tracing module and adaptive word selection module, both of which consult the model built using the components obtained by instantiating the concepts in the plant ontology. The system has been implemented and preliminary evaluation has been done successfully.

A Methodology of Collaborative Synthesis by Artificial Intelligence

Abstract: This project aims at establishing basic theories and fundamental methodologies for synthesis through knowledge systematization by exploiting the artificial intelligence technology, especially Ontological Engineering. The concrete objectives of the project include development several kinds of ontologies for knowledge systematization and development of a multi-agent collaborative synthesis framework. The formers mainly concerned with the domain-independent design object knowledge and the latter with domain-independent design process knowledge. This paper report on the intermediate results in the middle of the project term(five years) as well as on the future plan.

User modeling research in Japan

Abstract: This abstract overviews user modeling research in Japan. Not only modeling methods including learning algorithms but visualization for adaptive interaction and human modeling for plant operator have been extensively done. Modeling in learning support is also active. The activities are classified into the following five major categories.