人工知能学会論文誌
Online ISSN : 1346-8030
Print ISSN : 1346-0714
検索
OR
閲覧
検索
20 巻 , 6 号
選択された号の論文の11件中1~11を表示しています
    • |<
    • <
    • 1
    • >
    • >|
一般論文
論文
  • 鄭 耀陽, 五福 明夫
    20 巻 (2005) 6 号 p. 356-369
    公開日: 2005/08/23
    ジャーナル フリー
    This study investigates a technique to generate essential explanation for the reasoning process derived by the qualitative reasoning based on a functional model, in order to realize a mutual cooperation between a team of operators and an operator support system in an anomalous plant situation. This study applies a technique called function flow simplification to simplify a microscopic functional model into a macroscopic one. Then, this paper proposes a technique composed of the function flow simplification, simplification of functional model considering the purpose of a component and mapping the simplified functional model by the function flow simplification to a corresponding structure model to generate essential qualitative explanations. The applicability of the proposed technique is demonstrated by the explanation generation for effect propagations of counter actions for an anomaly in an oil refinery plant.
    抄録全体を表示
  • 浅田 洋平, 松尾 豊, 石塚 満
    20 巻 (2005) 6 号 p. 370-378
    公開日: 2005/08/23
    ジャーナル フリー
    Social networks, which describe relations among people or organizations as a network, have recently attracted attention. With the help of a social network, we can analyze the structure of a community and thereby promote efficient communications within it. We investigate the problem of extracting a network of researchers from the Web, to assist efficient cooperation among researchers. Our method uses a search engine to get the cooccurences of names of two researchers and calculates the streangth of the relation between them. Then we label the relation by analyzing the Web pages in which these two names cooccur. Research on social network extraction using search engines as ours, is attracting attention in Japan as well as abroad. However, the former approaches issue too many queries to search engines to extract a large-scale network. In this paper, we propose a method to filter superfluous queries and facilitates the extraction of large-scale networks. By this method we are able to extract a network of around 3000-nodes. Our experimental results show that the proposed method reduces the number of queries significantly while preserving the quality of the network as compared to former methods.
    抄録全体を表示
  • 鳥居 大祐, François Bousquet, 石田 亨
    20 巻 (2005) 6 号 p. 379-386
    公開日: 2005/08/30
    ジャーナル フリー
    Companion Modeling is a methodology of refining initial models for understanding reality through a role-playing game (RPG) and a multiagent simulation. In this research, we propose a novel agent model construction methodology in which classification learning is applied to the RPG log data in Companion Modeling. This methodology enables a systematic model construction that handles multi-parameters, independent of the modelers ability. There are three problems in applying classification learning to the RPG log data: 1) It is difficult to gather enough data for the number of features because the cost of gathering data is high. 2) Noise data can affect the learning results because the amount of data may be insufficient. 3) The learning results should be explained as a human decision making model and should be recognized by the expert as being the result that reflects reality. We realized an agent model construction system using the following two approaches: 1) Using a feature selction method, the feature subset that has the best prediction accuracy is identified. In this process, the important features chosen by the expert are always included. 2) The expert eliminates irrelevant features from the learning results after evaluating the learning model through a visualization of the results. Finally, using the RPG log data from the Companion Modeling of agricultural economics in northeastern Thailand, we confirm the capability of this methodology.
    抄録全体を表示
論文特集:オントロジーとその応用
論文
  • 兼岩 憲, 溝口 理一郎
    20 巻 (2005) 6 号 p. 387-395
    公開日: 2005/08/30
    ジャーナル フリー
    This paper presents several extensions of order-sorted logic based on the notion of property classification in formal ontology. The refined order-sorted language contains types (as rigid sorts), anti-rigid sorts, and unary predicates in order to distinctly express the following properties: substantial sorts, non-substantial sorts, and non-sortal properties. For many separated knowledge bases constructed using the logic, we propose an enriched reasoning mechanism such that each independent knowledge base can extract rigid property information from other knowledge bases (called rigid property derivation). Additionally, we classify (i) anti-rigid properties on the basis of their dependences on time, situation, and belief and (ii) non-sortal properties on the basis of the countability of the elements of the properties.
    抄録全体を表示
  • 桑原 教彰, 桑原 和宏, 安部 伸治, 須佐見 憲史, 安田 清
    20 巻 (2005) 6 号 p. 396-405
    公開日: 2005/09/08
    ジャーナル フリー
    Providing good home-based care to people with dementia is becoming an important issue as the size of the elderly population increases. One of the main problems in providing such care is that it must be constantly provided without interruption, and this puts a great burden on caregivers, who are often family members. Networked Interaction Therapy is the name we call our methods designed to relieve the stress of people suffering from dementia as well as that of their family members. This therapy aims to provide a system that interacts with people with dementia by utilizing various engaging stimuli. One such stimulus is a reminiscence video created from old photo albums, which is a promising way to hold a dementia sufferer's attention for a long time. In this paper, we present an authoring tool to assist in the production of a reminiscence video by using photo annotations. We conducted interviews with several video creators on how they used photo annotations such as date, title and subject of photos when they produced the reminiscence videos. According to the creators' comments, we have defined an ontology for representing the creators' knowledge of how to add visual effects to a reminiscence video. Subsequently, we developed an authoring tool that automatically produces a reminiscence video from the annotated photos. Subjective evaluation of the quality of reminiscence videos produced with our tool indicates that they give impressions similar to those produced by creators using conventional video editing software. The effectiveness of presenting such a video to people with dementia is also discussed.
    抄録全体を表示
  • 高井 貴子, 溝口 理一郎
    20 巻 (2005) 6 号 p. 406-416
    公開日: 2005/09/08
    ジャーナル フリー
    Signaling pathways are causal sequences of chemical reactions that end up by achieving cellular functions. Much information has been accumulated in databases for signaling pathways and ontologies have been built for sharing and reusing the information in community. However, we still have difficulties in consistent representation of signaling pathways. We still fail to have an intrinsic definition of a signaling pathway. We also fail to specify cooperative actions of molecules in dynamical construction of molecular complexes. We have developed Cell Signaling Networks Ontology (CSNO) based on device ontology to address theses difficulties. CSNO explicates that "signal" carries two kinds of information such as "identity" and "causality" of a signaling pathway. Identity is coded into "molecular recognition" predetermined in a cell, whereas causality is coded into "activity" processed along a pathway progressively and is essential for two-layered functions of cooperative molecules: to let activities flow (elementary function) and to control the flow of activities (mechanical function). CSNO defines roles of complex formation as dynamical construction of a device that performs a mechanical function generating cooperation in actions of molecules. Based on the definitions of signaling pathways, CSNO provides us with a base of integrative and consistent representation of signaling pathways, inducing a viewpoint in which not molecules but pathways are focused in the knowledge representation.
    抄録全体を表示
  • 高梨 勝敏, 佐藤 俊也, 原島 一郎
    20 巻 (2005) 6 号 p. 417-425
    公開日: 2005/09/13
    ジャーナル フリー
    We propose an idea called ``ontology reuse'' to develop ontology (or taxonomy) for information search more easily by its development cost reduction. For its feasibility demonstration, we developed a prototype system that converts taxonomy information of web contents to RDF format, collects it into a server and visualizes taxonomy infomation as a 2D map for editing itself and searching information. And our goal is to realize ``ontology circulation'', which means that ontology information flows freely and easily over the Internet and an intranet.
    抄録全体を表示
  • 川村 隆浩, 長谷川 哲夫, 大須賀 昭彦, Massimo Paolucci, Katia Sycara
    20 巻 (2005) 6 号 p. 426-436
    公開日: 2005/09/13
    ジャーナル フリー
    Experiment with public deployment of the semantic service matchmaker to a UDDI registry for a year is described in this paper. UDDI is a standard registry for Web Services, but, its search functionality has been limited to a keyword search. Therefore, we propose an enhancement of UDDI, called Matchmaker, that supports semantic service capability discovery. Since September 2003, we have deployed the Matchmaker in one of four official UDDI registries operated by NTT-Communications. In this paper, we first introduce the Matchmaker, and illustrate client tools which lower the threshold of use of semantics. Then, we evaluate this experiment by benchmarks in terms of its performance and functionality. Finally, we discuss user requirements obtained by two ways: questionnaire and observation of search behavior.
    抄録全体を表示
  • 星合 忠, 山根 康男, 津田 宏
    20 巻 (2005) 6 号 p. 437-447
    公開日: 2005/09/21
    ジャーナル フリー
    When ontology description data by different authors would become widespread in the world, we will be faced with the difficulties of the ontology alignment (OA) problem required for integration and interoperability of ontologies. The OA problem is the problem to find couples of semantically same classes / properties between two ontologies, and includes points of different naming of classes / properties, polysemous naming of classes / properties, different granularity of classes / properties, different hierarchical structures, and so on. We applied our semantic category matching (SCM) tool to the ontology alignment problems. Our approach found pairs of semantically corresponding categories from two different classification hierarchies such as Yahoo directory or library classification as UDC or NDC, based on natural language processing, similarity searching of huge vector spaces, and structural consistency analysis. We tackled problems of the EON2004 Ontology Alignment Contest. For examples, the Contest's random name problems (#201, #202) could not be solved using conventional character string resemblance techniques. However, when we applied SCM to these problems, the results showed that SCM had improved the accuracy as compared to the conventional method (F-measure: 0.021=>0.949, 0.021=>0.580), and exceeded the accuracy average in all problem areas by over 10 % as compared to conventional methods. Our team participated as a competitor in EON OA Contest and could obtain satisfactory results.
    抄録全体を表示
  • 藁谷 敏晴
    20 巻 (2005) 6 号 p. 448-460
    公開日: 2005/10/12
    ジャーナル フリー
    The main aim of this paper is to propose an appropriate logical system that is suitable to describe the notion of IS-A link as well as is-a link. The most important point to be realized is that those relations are not set thoretical ones. They connect two `general names' to construct a proposition, so that what is needed for proper descriptions of the relations in question is a theory of general names. It will be shown that is-a is a logical unit of axiomatically determined behaviour. The axiom concerning is-a relation was established by S. Lesniewski who named his theory of general names ontology. Today `ontology' has also become a common term for AI researchers. I intend to make it clear that there is a close connection between `ontology' used by Lesniewski and by AI researchers, even though they developed quite independently. I wish to stress that ontology created by Lesniewski is a system of syllogistic equipped with singular propositions and the theory of quantification. To make this point clear, I proposed a fragment of syllogism that I called MO(minimal ontology). This paper includes comments and examples articulating the logical power of ontology.
    抄録全体を表示
  • 砂川 英一, 古崎 晃司, 來村 徳信, 溝口 理一郎
    20 巻 (2005) 6 号 p. 461-472
    公開日: 2005/10/18
    ジャーナル フリー
    Based on some fundamental theories of ontology, we can treat a role concept as a concept which an entity plays in a context and discriminate it from a basic concept. Because a theory of role concepts makes a static policy for treatment of views and contexts related to conceptualization, discrimination of role concepts contributes effectively to management of instance models. In our research, we have developed an ontology building environment, which provides a framework for representation of role concepts and their characteristics. However, in the framework, role concepts are dealt with in a basic concept centered view and their definitions are scattered around in the respective related concepts which give the context of the roles. This is why users cannot easily represent relations among role concepts and grasp their whole image in an ontology. In this paper, as an extension of this framework, we present a framework for organizing role concepts in a hierarchy from role centered view. We investigate how to organize role concepts according to their contextual dependencies and focus on defining and organizing a role concept which depends on several contexts.
    抄録全体を表示
    • |<
    • <
    • 1
    • >
    • >|
feedback
Top