Journal of the Japanese Society for Artificial Intelligence
Online ISSN : 2435-8614
Print ISSN : 2188-2266
Volume 14, Issue 1
Displaying 1-36 of 36 articles from this issue
Print ISSN:0912-8085 until 2013
  • Kazuyo TANAKA
    Article type: Preface
    1999 Volume 14 Issue 1 Pages 1
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (109K)
  • Tsukasa HIRASHIMA
    Article type: Cover article
    1999 Volume 14 Issue 1 Pages 2
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (105K)
  • Tadahiko KUMAMOTO
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 3-10
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (889K)
  • Kenichi YOSHIDA
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 11-16
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (670K)
  • Tsukasa HIRASHIMA
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 17-24
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (934K)
  • Masanori SUGIMOTO
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 25-32
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (886K)
  • Yasuyuki SUMI
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 33-40
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (923K)
  • Hideyuki TAMURA
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 41-44
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (428K)
  • Syun TSUCHIYA
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 44-45
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (226K)
  • Toshio YOKOI
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 45-48
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (417K)
  • Susumu KUNIFUJI
    Article type: Cover article
    1999 Volume 14 Issue 1 Pages 49
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (103K)
  • Susumu KUNIFUJI
    Article type: Special issue
    1999 Volume 14 Issue 1 Pages 50-57
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (943K)
  • Kazushi NISHIMOTO, Kenji MASE, Ryohei NAKATSU
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 58-70
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    We have been developing a creativity support system called "AIDE, " which is equipped with various agents to stimulate creative group conversations. In this paper, we describe an autonomous information retrieval agent called "Conversationalist, " which is one of the agents of AIDE and is responsible for stimulating human divergent thinking. This agent analyzes the relationships among utterances and the structure of the topic in a conversation, and autonomously extracts various pieces of information relevant to the current conversation. Furthermore, we also show subjective experiments of AIDE applied to brainstorming sessions. From the results of the experiments, we confirmed that the agent is effective in stimulating human divergent thinking and in extracting more ideas from subjects, than in brainstorming sessions without the agent. Based on the results, we discuss what kind of information retrieval method is effective and when extracted pieces of information should be provided. Consequently, the following results are suggested : 1) when a conversation is active, the frequency of information provision by the agent should be rather low, and the relationship between the topic of the conversation and the pieces of information should not be so far, and 2) when a conversation is not active, i.e., is stagnate, or asynchronously executed, the frequency of information provision by the agent should be rather high, and the pieces of information should include some hidden relations with the topic of the conversation.

    Download PDF (1716K)
  • Kiyoshi NITTA, Lin TAO, Kazuo MISUE
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 71-81
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    An Idea Creation Support System (ICSS) requires a wide range of functions : including visualization, modeling, analysis and data management. There is no a functional boundary for an ICSS in practice and the implementations of the functions depends on the specific usage. Therefore, the dynamic extensibility and flexibility are the two challenge issues faced by ICSS developers. This paper presents an ICSS architecture which consists of a component model and a component management system. By centralizing signal running between components and the coherently integration of the component model and the componenet management system, the systems developed based on this architecture are highly flexible and dynamically extensible and can provide better support for the trial-and-error style idea creation process.

    Download PDF (1246K)
  • Yasuhiro YAMAMOTO, Shingo TAKADA, Kumiyo NAKAKOJI
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 82-92
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    The goal of this study is to design and build a computer system to support the basic cognitive activity of "writing" in a more natural and effective manner. The paper starts with a description of a writing process, followed by an overview of existing models on writing. Then, the notion of "Representational Talkback" is proposed as an important aspect in supporting collage-style writing. Representational Talkback is defined as "feedback from externally represented artifacts." The ART (Amplifying Representational Talkback) system is implemented based on this notion, focusing on the role of meta-comments in writing. The goal of the system is twofold : (1) to support collage-style writing of a document, and (2) to observe how people "write" using ART. The paper concludes with a discussion of the result of a study on how people "write" using ART with an eye towards extending the notion to other types of cognitive activities.

    Download PDF (1287K)
  • Yasufumi TAKAMA, Mitsuru ISHIZUKA
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 93-101
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    Recent rapid growth of information environment such as the Internet makes it easy for us to get vast information. On the other hand, "information overflow" is becoming a serious problem. That is, available information resources are too much for human to utilize, and sometimes confuse human's activity. In this paper, we propose a new mechanism for discovering a similarity among documents. This mechanism is called "Fish Eye Matching", which generates a feature vector from a text dynamically based on a concept structure derived from an electronic dictionary such as EDR. Using this concept structure, this mechanism can use semantic relations among words, which have not been considered in existing feature vector paradigms, and reorganize information space dynamically according to users' interests. This paper also presents a prototype system for supporting information organization process based on this Fish Eye Matching method.

    Download PDF (936K)
  • Yasuyuki SUMI, Koichi HORI, Setsuo OHSUGA
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 102-110
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    This paper describes a system for supporting the construction of requirement models, which are initial computable models representing users' requirements in system design. The proposed system principally consists of two components, a tool for aiding the formation of requirement concepts by visualizing a user's thought space, and a knowledge-based system which automatically assembles the ascertained requirement concepts into a requirement model. The system extracts reusable components of a requirement model, corresponding to the users' abstract requirement concept, from a store of similar past cases. The components are then automatically arranged using heuristic reasoning. By using the system, users can make their requirement concepts more mature, and simultaneously get computable requirement models as by-products.

    Download PDF (948K)
  • Chie KADOWAKI, Tomohiro KOKOGAWA, Toshihiko YAMAKAMI, Keizo SUGITA, Su ...
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 111-121
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    Information sharing within an organization using computer networks is becoming more important each day in line with the expansion of the Internet. We are interested in the issue of information sharing on computer networks within an organization. Therefore, we examine user behavior in terms of information acquisition using computer networks. The results of this study demonstrate the importance of organizational information sharing. The study of retrieval and referential logs recorded during a twelve-month experiment of a prototype know-how management system highlights the importance of three factors in organizational information sharing : interest in shared information use, curiosity about shared information, and clues on how to use shared information. Based on our observations, we propose a concept called "Information Acquisition Awareness" that supports organizational information acquisition by addressing these factors. The support functions related to these three factors are : (1) stimulating the user's latent interests in shared information use, (2) arousing the user's curiosity about shared information, and (3) giving opportunities to acquire clues for using shared information. These three functions are integrated into a support system, called GGG, by extracting useful information automatically from shared information. The effectiveness of GGG is shown by the evaluation. The results show that the proposed support functions have significant effects on promoting organizational information sharing when users notice the presence and the value of shared information, which had previously lain idle in the system.

    Download PDF (1205K)
  • Hajime KIMURA, Shigenobu KOBAYASHI
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 122-130
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    Many previous works in reinforcement learning (RL) are limited to Markov decision processes (MDPs). However, a great many real-world applications do not satisfy this assumption. RL tasks of real world can be characterized by two difficulties : Function approximation and hidden state problems. For large and continuous state or action space, the agent has to incorporate some form of generalization. One way to do it is to use general function approximators to represent value functions or control policies. Hidden state problems, which can be represented by partially observable MDPs (POMDPs), arise in the case that the RL agent cannot observe the state of the environment perfectly owing to noisy or insufficient sensors, partial information, etc. We have presented a RL algorithm in POMDPs, that is based on a stochastic gradient ascent. It uses function approximator to represent a stochastic policy, and updates the policy parameters. We apply the algorithm to a robot control problem, and show the features in comparison with Q-learning or Jaakkola's method. The results shows the algorithm is very robust under the conditions that the agent is restricted to computationally very poor resources.

    Download PDF (1030K)
  • Einoshin SUZUKI
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 139-147
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    This paper presents an evaluation method for discovering probabilistic if-then rules with high reliability from data sets. The discovery of probabilistic if-then rules, each of which is a restricted form of a characteristic production rule, is well motivated by various useful applications such as the semantic query optimization and the automatic development of a knowledge-base. In a discovery algorithm, a production rule is evaluated according to its generality and its accuracy since these are widely accepted as criteria in inductive learning. Here, reliability evaluation for these criteria is mandatory in distinguishing reliable rules from unreliable patterns without annoying the users. However, previous discovery approaches for characteristic rules have either ignored the reliability evaluation or have only evaluated the reliability of generality. Consequently, they tend to discover a huge number of rules, some of which are unreliable in their accuracies. In order to circumvent these difficulties we propose an approach based on a simultaneous estimation. Our approach discovers, based on the normal approximations of the multinomial distributions, the rules which exceed the pre-specified thresholds for generality and accuracy with high reliability. A novel pruning method is employed for improving the time efficiency without changing the discovery outcome. The proposed approach has been validated experimentally using 21 benchmark data sets in the machine learning community.

    Download PDF (857K)
  • Kazuteru MIYAZAKI, Sachiyo ARAI, Shigenobu KOBAYASHI
    Article type: Technical paper
    1999 Volume 14 Issue 1 Pages 148-156
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    Partially Observable Markov Decision Process (POMDP) is a representative class of non-Markovian environments, where agents sense different environmental states as the same sensory input. We recognize that full implementation of POMDPs must overcome two deceptive problems. We call confusion of state values a Type 1 deceptive problem and indistinction of rational and irrational rules a Type 2 deceptive problem. The Type 1 problem deceives Q-learning, the most widely-used method in which state values are estimated. Though Profit Sharing that satisfies Rationality Theorem [Miyazaki 94] is not deceived by Type 1 problem, it cannot overcome a Type 2 problem. A current approach to POMDPs is classified into two types. One is the memory-based approach that uses histories of sensor-action pairs to divide partially observable states. The other is to use stochastic policy where the agent selects action stochastically to escape from partially observable states. The memory-based approach needs numerous memories to store histories of sensor-action pairs. Stochastic policy may generate unnecessary actions to acquire rewards. In this paper, we propose a new approach to POMDPs. For the subclass environment that does not need stochastic policy, we consider to learn a deterministic rational policy to avoid all states that manifest a Type 2 problem. We claim that the weight as an evaluation factor of a rule has the possibility to derive an irrational policy due to Type 2 problem. Therefore, no weight is used to make a rational policy. We propose the Rational Policy Making algorithm (RPM) that can learn a rational policy by direct acquirement of rational rules from that rule's definition. RPM is applied to maze environments. We show that RPM can learn the most stable rational policy in comparison with other methods.

    Download PDF (889K)
  • Takahiro HAYASHI, Haruhiko KIMURA, Sadaki HIROSE, Shigeki HIROBAYASHI
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 166-173
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS

    Production systems are established methods for encoding knowledge in expert systems. In production systems, it's an important problem how to deal with expensive productions. Direct-match algorithm that is an algorithm for dealing with expensive productions. But this algorithm needs large memory space for getting maximum efficiency. In this paper, we propose a new Direct-match algorithm dealing with expensive productions without using large memory space.

    Download PDF (774K)
  • [in Japanese]
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 174-176
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (222K)
  • [in Japanese]
    Article type: Other
    1999 Volume 14 Issue 1 Pages 177-178
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (228K)
  • [in Japanese], [in Japanese], [in Japanese], [in Japanese]
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 179-181
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (283K)
  • [in Japanese]
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 182
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (114K)
  • [in Japanese]
    Article type: Corner article
    1999 Volume 14 Issue 1 Pages 183
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (113K)
  • Article type: Activity report
    1999 Volume 14 Issue 1 Pages 184-188
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (361K)
  • Article type: Activity report
    1999 Volume 14 Issue 1 Pages 189-190
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (200K)
  • Article type: Activity report
    1999 Volume 14 Issue 1 Pages 191-197
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (452K)
  • Article type: Activity report
    1999 Volume 14 Issue 1 Pages 198
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (104K)
  • Article type: Activity report
    1999 Volume 14 Issue 1 Pages b001-b012
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (1054K)
  • Article type: Cover page
    1999 Volume 14 Issue 1 Pages c001
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (49K)
  • Article type: Cover page
    1999 Volume 14 Issue 1 Pages c001_2
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (108K)
  • Article type: Table of contents
    1999 Volume 14 Issue 1 Pages i001
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (92K)
  • Article type: Table of contents
    1999 Volume 14 Issue 1 Pages i001_2
    Published: January 01, 1999
    Released on J-STAGE: September 29, 2020
    MAGAZINE FREE ACCESS
    Download PDF (98K)
feedback
Top