著者

前野 義晴 森永 聡 松島 宏和 天谷 健一

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.27, no.6, pp.338-345, 2012 (Released:2012-10-02)

参考文献数

21

被引用文献数

1 2

---

This study presents a computer simulation model to analyze the risk of transmission of financial distress in a bank credit network and the knock-on defaults of banks. The impact is quantified, which is imposed on the number of defaults by the topology of the bank credit network, the balance sheet of banks including equity capital ratio, and the capital surcharge on big banks.

著者

増田 壮志 古崎 晃司

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.32, no.2, pp.E-G71_1-10, 2017-03-01 (Released:2017-03-01)

参考文献数

17

---

Ontologies are currently constructed in various fields, such as life sciences, medical information, and sustainability science. These ontologies are used as knowledge bases and knowledge models for application systems. However, it is difficult to build high quality ontologies due to the necessity of having both knowledge of ontology and expertise in the target domain. Therefore, ontology construction and maintenance costs considerable time and effort. To reduce such costs, we developed an ontology refinement support method. To test and confirm this refinement method, we focused on the guideline for building well-organized ontologies that“ Each subclass of a super class is distinguished by the values of exactly one attribute of the super class. ”Then, we discovered that there is a similarity between is-a hierarchies when an ontology is built following this guideline and made the hypothesis that, if subclasses are not classified by one attribute, there are consistency errors in the ontology that can be automatically fixed by a comparison method of is-a hierarchies. To test this hypothesis, we conducted an experiment to evaluate the refinement method. We asked nine experienced evaluators to build the ontology and used 150 refinement proposals. As a result, we found that at most 90% of the refinement candidates could be further refined and that at most approximately 50% of the refinement proposals are appropriate to apply to ontologies.

著者

山下 諒 朴 炳宣 松下 光範

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.32, no.1, pp.WII-D_1-11, 2017-01-06 (Released:2017-01-20)

参考文献数

29

---

The purpose of this research is supporting information access based on the contents of comic books. To meet this purpose, it is necessary to obtain information related to the story and the characters of a comic. We propose a method to extract information from reviews on the Web by using term frequency-inversed document frequency (TFIDF) method and hierarchical Latent Dirichlet Allocation (hLDA) method, which intends to solve the problem. By using these methods, we build a prototype system for exploratory comic search. We conducted a user study to observe how a participant use the system. The user study showed that the system successfully supported the participants to find interesting unread comics.

著者

大谷 世紀 山西 良典 岩堀 祐之

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.32, no.1, pp.WII-N_1-10, 2017-01-06 (Released:2017-01-20)

参考文献数

17

被引用文献数

3

---

Image database is one of the important research topics in image recognition. A manual collection of images would causes biased collection and a lot of human efforts. Huge image database where varied and many unbiased objects are stored is required for the image learning. Recent researches for image database try to automatically/semiautomatically generate image database. Here, it is important to remove noise images and this paper proposes an automatic generation method of the Web image database. The proposed method uses noise image removal by visual feature and semantic feature in a hybrid. However, which type of features, and how to combine the two types of feature are not clear and should be investigated. In this paper, six kinds of noise image detection method are prepared: The method using visual feature, the method using semantic feature, two methods using both features in parallel and two methods using both features in serial. Through the comparison in experiments, it was confirmed that the method using both visual and semantic features in parallel focusing on noise images showed over 82% Precision values,76% Recall values and 77% Fmeasure values in average. Also, the usability of the generated database for image recognition was confirmed through the experiments; It was equal to or higher than the human-made database. It was confirmed that the proposed method constructed precise image database full-automatically.

著者

松村 真宏 大澤 幸生 石塚 満

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.17, no.3, pp.259-267, 2002 (Released:2002-04-25)

参考文献数

14

被引用文献数

13 36

---

Business people, especially marketing researchers, are keen to understand peoples' potential sense of value to create fascinating topics stimulating peoples' interest. In this paper, we aim at finding influential people, comments, and terms contributing the discovery of such topics. For this purpose, we propose an Influence Diffusion Model in text-based communication, where the influence of people, comments, and terms are defined as the degree of text-based relevance of messages. We apply this model to Bulletin Board Service(BBS) on the Internet, and present our discoveries on experimental evaluations.

著者

松原 仁 田中 久美子 Frank Ian 田所 諭

出版者

The Japanese Society for Artificial Intelligence

雑誌

人工知能学会論文誌 = Transactions of the Japanese Society for Artificial Intelligence : AI (ISSN:13460714)

巻号頁・発行日

vol.17, pp.177-180, 2002-11-01

被引用文献数

4 2

---

We address the problem of information flow in disaster relief scenarios by presenting an architecture for generating natural language dialogue between large numbers of agents. This architecture is the first step towards real-time support systems for relief workers and their controllers. Our work demonstrates how natural generation techniques from the MIKE commentary system for RoboCup soccer can be carried over to that of RoboCup Rescue. Thanks to this background, the initial product of our research is a system that explains a RoboCup Rescue simulation not to the agents in the domain themselves but to a watching audience. This "commentary" is produced by recreating the actual dialogues most likely be occurring in the domain: walkie-talkie-conversations.

著者

菊地 剛正 國上 真章 山田 隆志 高橋 大志 寺野 隆雄

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.31, no.6, pp.AG-G_1-11, 2016-11-01 (Released:2016-11-02)

参考文献数

33

被引用文献数

2

---

Since systemic risk in recent financial crisis has received attention, a wide variety of studies on how financial regulations should be reformed to control such risk have been progressed. Among these, mathematical and computational studies have analyzed the way how the borrowing and lending banks and the borrowers go bankrupt in the chain via interbanking network. This study deals with a chain of bankruptcies of financial institutions endogenously caused by deterioration of their financial situations from the changes in prices of risky assets by focusing on the macro-level collapse caused by the shock stemming from the general market risk factors addressed in theoretical studies after the crisis in 2008. For this purpose, the authors develop an agent-based simulation platform and then examine how current systemic management regulations affect corresponding bankruptcies. The main findings are as follows: First, pertinent management regulations are dependent on the market environment. Second, some combination of management regulations may increase the possibilities of bankruptcies due to more sensitivity toward the market change.

著者

伊藤 一之 松野 文俊

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.16, no.6, pp.510-520, 2001 (Released:2002-02-28)

参考文献数

19

被引用文献数

2 7

---

Reinforcement learning has recently received much attention as a learning method for complicated systems, e.g., robot systems. It does not need prior knowledge and has higher capability of reactive and adaptive behaviors. However increase in dimensionality of the action-state space makes it diffcult to accomplish learning. The applicability of the existing reinforcement learning algorithms are effective for simple tasks with relatively small action-state space. In this paper, we propose a new reinforcement learning algorithm: “Q-learning with Dynamic Structuring of Exploration Space Based on Genetic Algorithm ”. The algorithm is applicable to systems with high dimensional action and interior state spaces, for example a robot with many redundant degrees of freedom. To demonstrate the effectiveness of the proposed algorithm simulations of obstacle avoidance by a 50 links manipulator have been carried out. It is shown that effective behavior can be learned by using the proposed algorithm.

著者

Hirahara Yuki Toriumi Fujio Sugawara Toshiharu

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

pp.JWEIN-K, (Released:2016-01-28)

参考文献数

17

---

We propose an SNS-norms game to model behavioral strategies in social networking services (SNSs) and investigate the conditions required for the evolution of cooperation-dominant situations. SNSs such as Facebook and Google+ are indispensable social media for a variety of social communications ranging from personal chats to business and political campaigns, but we do not yet fully understand why they thrive and whether these currently popular SNSs will remain in the future. A number of studies have attempted to understand the conditions or mechanisms that keep social media thriving by using a meta-rewards game that is the dual form of a public goods game or by analyzing user roles. However, the meta-rewards game does not take into account the unique characteristics of current SNSs. Hence, in this work we propose an SNS-norms game that is an extension of Axelrod's metanorms game, similar to meta-rewards games, but that considers the cost of commenting on an article and who is most likely to respond to it. We then experimentally investigated the conditions for a cooperation-dominant situation, by which we mean many users continuing to post articles on an SNS. Our results indicate that relatively large rewards compared to the cost of posting articles and comments are required to evolve cooperation-dominant situations, but optional responses with lower cost, such as ``Like!'' buttons, facilitate the evolution. This phenomenon is of interest because it is quite different from those shown in previous studies using meta-rewards games. We also confirmed the same phenomenon in an additional experiment using a network structure extracted from real-world SNS data.

著者

中道 大介 西尾 修一

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.31, no.2, pp.H-F81_1-10, 2016-03-01 (Released:2016-02-18)

参考文献数

12

被引用文献数

5

---

Teleoperation enables us to act in remote location through operated entities such as robots or virtual agents. This advantage allows us to work in places dangerous for humans or places not designed for humans such as in volcano disaster site or in narrow maintenance pipes. However, teleoperation also has a weakness, namely, several gaps (operation interface, environment, appearance, and intentionality) among ourselves and the teleoperated entities in remote. As teleoperated robots own physical bodies different from us, teleoperation requires special interfacing systems that are usually not so intuitive. Such a system requires rather long period of training for one to become familiar with it. One possible solution for this issue is to implement semi-autonomous teleoperation (SAT) facility which combines manual operation and autonomous action. With SAT, we can teleoperate remote entities in a way suitable to the teleoperated entity and to the situation in remote with the help of autonomous action. However, there is a concern that this autonomous part of SAT may decrease operators' feeling of agency and consequently, operators may lose concentration and become less efficient. To deal with this issue, we made a hypothesis that if an autonomously generated action is sufficiently suitable to the situation, operators would feel their own agency to the generated action. In this paper, we examined this hypothesis through an experiment where participants joined a conversation using a teleoperated android robot. Here we focused on autonomous generation of nodding act and evaluated operators' agency toward the generated actions. Participants listened to a speech through the teleoperated robot with different degree of autonomous motion generation. After a series of listening, they watched video recordings taken from the speaker's view that showed the teleoperated robot in action, and evaluated their agency toward the action and the appropriateness of robot's motion toward the speaker. As a result, we found that as timing of nodding become appropriate to the conversation context, the operators kept agency to robot's motion, even when the automatically generated motion and the operator's own motion were mixed. Furthermore, the robot motion which was automatically generated was evaluated more appropriate to the conversation than the operator's own motion. In conclusion, when using semi-autonomous teleoperation, if the autonomous action is appropriate to the situation, operators are able to keep agency to operated entities' acts.

著者

早矢仕 晃章 大澤 幸生

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.31, no.5, pp.A-G15_1-9, 2016-09-15 (Released:2016-08-15)

参考文献数

16

被引用文献数

6

---

The potential desire of companies for creating values by combining data from different domains has been increased. In order to lead data-driven innovations, a market of data is expected to enhance this combination and data exchange through the communication among stakeholders. Innovators Marketplace on Data Jackets (IMDJ) is a gamified workshop for discovering the value of data by discussing the combination of Data Jackets, which supports creativity toward innovations and activates a market of data. A Data Jacket is meta-data, i.e., a summary of a dataset. Even if the data is not open, a Data Jacket enables participants to consider the latent value of datasets through creative communication. In this study, we discuss a system for structuring and reusing knowledge of data utilization, which are created in the workshops of IMDJ. By modeling and structuring knowledge not only with datasets, but also with solutions or requirements, it is expected to be possible to retrieve important information about solving problems. By implementing structured knowledge of data utilization using RDF (Resource Description Framework) and designing the interface for extracting accurate information for users, we propose the retrieval system named Data Jacket Store, and evaluate the performance.

著者

小林 重信

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.24, no.1, pp.147-162, 2009 (Released:2009-01-06)

参考文献数

37

被引用文献数

43 63

---

Real-coded genetic algorithms (RCGA) are expected to solve efficiently real parameter optimization problems of multimodality, parameter dependency, and ill-scale. Multi-parental crossovers such as the simplex crossover (SPX) and the UNDX-m as extensions of the unimodal normal distribution crossove (UNDX) show relatively good performance for RCGA. The minimal generation gap (MGG) is used widely as a generation alternation model for RCGA. However, the MGG is not suited for multi-parental crossovers. Both the SPX and the UNDX-m have their own drawbacks respectively. Therefore, RCGA composed of them cannot be applied to highly dimensional problems, because their hidden faults appear. This paper presents a new and robust faramework for RCGA. First, we propose a generation alternation model called JGG (just generation gap) suited for multi-parental crossovers. The JGG replaces parents with children completely every generation. To solve the asymmetry and bias of children distribution generated by the SPX and the UNDX-m, an enhanced SPX (e-SPX) and an enhanced UNDX (e-UNDX) are proposed. Moreover, we propose a crossover called REX(φ,n+k) as a generlization of the e-UNDX, where φ and n+k denote some probability distribution and the number of parents respectively. A concept of the globally descent direction (GDD) is introduced to handle the situations where the population does not cover any optimum. The GDD can be used under the big valley structure. Then, we propose REXstar as an extention of the REX(φ,n+k) that can generate children to the GDD efficiently. Several experiments show excellent performance and robustness of the REXstar. Finally, the future work is discussed.

著者

岡田 正人 金盛 克俊 青木 伸 大和田 勇人

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.29, no.1, pp.194-200, 2014-01-05 (Released:2014-01-07)

参考文献数

23

---

This paper presents a high performance virtual screening method for drug design based on machine learning. In drug discovery with computers, drug designers often use docking softwares. They decide the docking between the compound and the protein with the result of docking software, structure of the compound, and any information of the compound. Currently, the performance of docking software is not high. This paper shows the machine learning method which uses the experiential knowledge of pharmaceutical researchers. This method calculates the docking possibility of compounds with high performance based on the results of the docking software and chemical information of compounds. The experiment shows our method have high-accuracy as 98.4 % and excellent ROC curve.

著者

小林 一樹 船越 孝太郎 小松 孝徳 山田 誠二 中野 幹生

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.30, no.4, pp.604-612, 2015-07-01 (Released:2015-05-21)

参考文献数

30

被引用文献数

2

---

In this paper, we describe an investigation into users' experiences of a simple talking robot with back-channel feedbacks that is designed based on an artificial subtle expression (ASE). In the experiments with participants, they are divided into six conditions based on an expression factor (three levels; human-like speech, blinking light, and beeping sound) and a timing decision method factor (two levels; a linguistic method and an acoustic method) for investigating participants' impressions on the dialogue experience. We developed an electric pedestal to show the blinking expression, on which a simple cubic robot was fixed. Participants engaged in a task of explaining a cooking procedure with a spoken dialogue system coupled with the robot on the pedestal. The robots responded to them by making the back-channel feedbacks in accordance with the expression factor. The results of questionnaire analyses suggested that the ASE-based expressions of back-channel feedback provide positive experiences for users.

著者

中井 淳一 筑紫 晴久

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.30, no.6, pp.791-801, 2015-11-01 (Released:2016-04-06)

参考文献数

32

---

The structure of the control software is often complicated. The reason is that it's created by multiple developers and added the functions later. Thus, many software developers want technology to reconstruct decomposing the structure of the software. However, current situation, the grouping of the control software is dependent on the experience and the sense of the skilled person. In addition, the opportunity of the review is very small, generic clustering algorithm has not yet been established. Existing clustering algorithm is not considered to apply to the control software. The problem is the inter-group feedback increase and group size not adjustable. This is because, when inter-group feedback is large, it is difficult to understand the control software, rework increases when division of labor. Also, when there is feedback between the pre-process and the next process at the black box testing, the calculation result of the next step also affects the pre-process, the test process is increased considerably. In this study, we tried the application of the clustering algorithm of graph theory to structure organize the control software. Using a genetic algorithm, and by forming groups based on modularity, while groups of closely related ones, and aid the formation of a grouping less inter-group feedback. Furthermore, it is easy to understand, and for ease of testing, adjustment of the number of groups to a size suitable for control software also aims to be a possibility.

著者

仲田 圭佑 村田 剛志

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.30, no.1, pp.96-101, 2015-01-06 (Released:2015-01-06)

参考文献数

12

被引用文献数

1

---

Recent development of information technology and rise of social media enable us to access massive data. Large scale data such as hyperlink structure in WWW and friendship information in social media can be represented as networks based on graph theory. For analyzing such data, many methods have been proposed. Among them, the methods called community detection have advantages that they can make networks simple and easy to understand. However, most of them had not considered the background knowledge of data, thus some methods called constrained community detection which take such background knowledge into consideration have been proposed. Constrained community detection methods show robust performance on noisy data due to its background knowledge. In particular, constrained Hamiltonian-based community detection methods have advantages such as flexibility of output results. The Hamiltonian, energy in statistical mechanics, can be theoretically considered as a generalization of the Newman's modularity. In this paper, we propose a method for accelerating constrained community detection based on Hamiltonian. Our proposed method is a variant of Blondel's Louvain method which is known for its computational efficiency. We experimentally show that the proposed method is superior to the existing method based on simulated annealing in terms of computational efficiency, and its accuracy is as well as the existing method under the same conditions. Our method enables us to perform constrained community detection for larger networks compared with the existing method.

著者

佐藤 晃矢 岡 瑞起 橋本 康弘 加藤 和彦

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.30, no.5, pp.667-674, 2015-09-01 (Released:2015-08-27)

参考文献数

19

---

Social Tagging System (STS) which is one of the content management techniques is widely adopted in the online content sharing service. Using STS, users can give any strings (tags) to contents as annotations. It is important to know the usage of tag statistics for accomplishing an effective database design and the information navigation. The frequency of tag usage as well as their dynamics are similar to the ones found in the natural language. It is possible to reproduce the branching process of the tag dynamics using a classical model called Yule-Simon process. Another characteristic aspect of tags is the tag co-occurrence generated from the simultaneous use of tags. Using the tag co-occurrence, STS is able to reconstitute the hierarchy of tags, and recommend the tag which is probably used next. However, Yule-Simon process does not consider the tag co-occurrence and thus how the tag co-occurrence is generated from the model like Yule-Simon has not been addressed yet. In this paper, we propose to expand the Yule-Simon process to model the tag co-occurrence. From the point of view of network hierarchy, we confirm the similarity in the structure of the tag co-occurrence with the empirical data obtained from a social network service called ‘RoomClip’. The present result suggested that this simple model like extended Yule-Simon process generates the tag co-occurrence feature.

著者

古崎 晃司 來村 徳信 溝口 理一郎

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

pp.LOD-214, (Released:2015-12-25)

参考文献数

24

被引用文献数

1

---

Biomimetics contributes to innovative engineering by imitating the models, systems, and elements of nature. Well-known examples of biomimetics include paint and cleaning technologies that imitate the water repellency of the lotus, adhesive tapes that imitate the adhesiveness of gecko feet, and high-speed swimsuits that imitate the low resistance of a shark’s skin. These results integrate studies on the biological mechanisms of organisms with engineering technologies to develop new materials. Facilitating such biomimetics-based innovations requires integrating knowledge, data, requirements, and viewpoints across different domains. Researchers and engineers need to develop a biomimetics database to assist them in achieving this goal. Because ontologies clarify concepts that appear in target domains, we assume that it is important to develop a biomimetics ontology that contributes to improvement of knowledge interoperability between the biology and engineering domains. Furthermore, linked data technologies are very effective for integrating a database with existing biological diversity databases. On the basis of these observations, we developed a biomimetics ontology and keyword exploration tool based on linked data techniques. The tool allows users to find important keywords for retrieving meaningful knowledge from various biomimetics databases. Such a technique could support idea creation by users based on a biomimetics ontology. This paper describes a prototype of our proposed biomimetics ontology and keyword exploration tool.

著者

Katsuhiko Hayashi Jun Suzuki Masaaki Nagata

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

pp.J-F83, (Released:2015-12-17)

参考文献数

24

被引用文献数

1

---

The spinal tree adjoining grammar (TAG) parsing model of [Carreras 08] achieves the current state-of-the-art constituent parsing accuracy on the commonly used English Penn Treebank evaluation setting. Unfortunately, the model has the serious drawback of low parsing efficiency since its Eisner-CKY style parsing algorithm needs O(n4) computation time for input length n. This paper investigates a more practical solution and presents a beam search shift-reduce algorithm for spinal TAG parsing. Since the algorithm works in O(bn) (b is beam width), it can be expected to provide a significant improvement in parsing speed. However, to achieve faster parsing, it needs to prune a large number of candidates in an exponentially large search space and often suffers from severe search errors. In fact, our experiments show that the basic beam search shift-reduce parser does not work well for spinal TAGs. To alleviate this problem, we extend the proposed shift-reduce algorithm with two techniques: Dynamic Programming of [Huang 10a] and Supertagging. The proposed extended parsing algorithm is about 8 times faster than the Berkeley parser, which is well-known to be fast constituent parsing software, while offering state-of-the-art performance. Moreover, we conduct experiments on the Keyaki Treebank for Japanese to show that the good performance of our proposed parser is language-independent.

著者

杉山 弘晃 目黒 豊美 東中 竜一郎

出版者

一般社団法人 人工知能学会

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

pp.DSF-518, (Released:2015-12-15)

参考文献数

15

---

In conversational dialogue, a talker sometimes asks questions that relate to the other talker's personality, such as his/her favorites and experiences. This behavior also appears in conversational dialogues with a dialogue system; therefore, the system should be developed so that it responds to this kind of questions. Previous systems realized this function by creating question-answer pairs by hand. However, there is no work that examines the coverage of the created question-answer pairs over real conversations. This study analyzes a huge amount of question-answer pairs created by many question-generators, with one answer-generator for each character. Our analysis shows that 41% of personality questions that appeared in real conversations are covered by the created pairs. We also investigated the types of questions that are frequently asked.