著者

市川 類

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.37, no.3, pp.IDS-A_1-9, 2022-05-01 (Released:2022-05-01)

参考文献数

32

---

AI technology brings major transformation to the human society, while it may conflict with the values (ethics) of the human society, depending on how it is used. With this background, a number of “AI principles” have been developed as a code to AI ethics by major countries and institutions.Intelligent dialogue systems (IDSs), as a part of AI technologies, will play an important role as an interface of "human-machine (AI system)" in the future, and therefore, developers and users of IDSs need to be aware and manage the risk of ethical aspects of IDSs, including those related to "human-machine" relationship.The purpose of this paper is to clarify the possible ethical risks (including legal risks) of IDSs through analysis of AI regulation in Europe, which leads the discussion on AI ethics in the world, with its reference of its social acceptance and cultural background regarding “human-machine” relationship.For that purpose, this paper first shows the direction of future technological development of the IDSs, and then identifies the characteristics of AI principles of Europe, which may be affected by its cultural background related to the "human-machine" relationship. Then, through analysis of the recently proposed European AI Act with considering future technological development of IDSs and its characteristics, this paper clarifies the possible ethical risks for the future development and practical application of IDSs are not only those related to shared human rights such as fairness/non-discrimination and privacy but also those related to the cultural differences on the view of “human-machine” relationship.

著者

中林 一貴 益井 博史 谷口 忠大

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.35, no.5, pp.G-K31_1-10, 2020-09-01 (Released:2020-09-01)

参考文献数

19

---

Communication-field mechanism design is to design a mechanism including rules and incentives to indirectly control a group of people having communication, e.g., discussion, debate, meeting, and consultation. A communication-field mechanism is expected to give constraints to the actual communications. We hypothesized that such constraints are beneficial for the application of technologies based on artificial intelligence. In this paper, we evaluate this concept by taking an automatic speech recognition system and dealing rights to speak (DRS) as an example of proof of concept. An experiment shows that the simple introduction of DRS improves the performance of speech recognition.

著者

田村 浩一郎 上野山 勝也 飯塚 修平 松尾 豊

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.33, no.1, pp.A-H51_1-11, 2018-01-01 (Released:2018-01-05)

参考文献数

20

被引用文献数

4

---

In order to encourage individual asset flow into the Japanese market through long-term investments, it is important to evaluate stock values of companies because stock prices of companies are determined not only by internal values, which are independent of other companies, but also by market fundamentalism. However, there are few studies conducted in this area in the machine learning community, while there are many studies about prediction of stock price trends. These studies use a single factor approach (such as textual or numerical) and focus on internal values only. We propose a model where we combine two major financial approaches to evaluate stock values: technical analysis and fundamental analysis. The technical analysis is conducted using Long-Short Term Memory and technical indexes as input data. On the other hand, the fundamental analysis is conducted transversely and relatively by creating a program which can retrieve data on financial statements of all listed companies in Japan and put them into a database. From the experiments, compared to single technical analysis proposed model’s accuracy in classification was 11.92% more accurate and the relative error of regression was 3.77% smaller on average. In addition, compared to single factor approaches the accuracy in classification was 6.16% more accurate and the relative error of regression was 3.22% smaller on average. The proposed model has the potential to be combined with other prediction methods, such as textual approaches or even traditional financial approaches, which would improve accuracy and increase practicality of this model.

著者

鳥海 不二夫 山本 仁志 諏訪 博彦 岡田 勇 和泉 潔 橋本 康弘

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.25, no.1, pp.78-89, 2010 (Released:2010-01-06)

参考文献数

17

被引用文献数

3 9

---

In recent years, application of Social Networking Services (SNS) and Blogs are growing as new communication tools on the Internet. Several large-scale SNS sites are prospering; meanwhile, many sites with relatively small scale are offering services. Such small-scale SNSs realize small-group isolated type of communication while neither mixi nor MySpace can do that. However, the studies on SNS are almost about particular large-scale SNSs and cannot analyze whether their results apply for general features or for special characteristics on the SNSs. From the point of view of comparison analysis on SNS, comparison with just several types of those cannot reach a statistically significant level. We analyze many SNS sites with the aim of classifying them by using some approaches. Our paper classifies 50,000 sites for small-scale SNSs and gives their features from the points of network structure, patterns of communication, and growth rate of SNS. The result of analysis for network structure shows that many SNS sites have small-world attribute with short path lengths and high coefficients of their cluster. Distribution of degrees of the SNS sites is close to power law. This result indicates the small-scale SNS sites raise the percentage of users with many friends than mixi. According to the analysis of their coefficients of assortativity, those SNS sites have negative values of assortativity, and that means users with high degree tend to connect users with small degree. Next, we analyze the patterns of user communication. A friend network of SNS is explicit while users' communication behaviors are defined as an implicit network. What kind of relationships do these networks have? To address this question, we obtain some characteristics of users' communication structure and activation patterns of users on the SNS sites. By using new indexes, friend aggregation rate and friend coverage rate, we show that SNS sites with high value of friend coverage rate activate diary postings and their comments. Besides, they become activated when hub users with high degree do not behave actively on the sites with high value of friend aggregation rate and high value of friend coverage rate. On the other hand, activation emerges when hub users behave actively on the sites with low value of friend aggregation rate and high value of friend coverage rate. Finally, we observe SNS sites which are increasing the number of users considerably, from the viewpoint of network structure, and extract characteristics of high growth SNS sites. As a result of discrimination on the basis of the decision tree analysis, we can recognize the high growth SNS sites with a high degree of accuracy. Besides, this approach suggests mixi and the other small-scale SNS sites have different character trait.

著者

渡辺 知恵美 中村 聡史

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.340-352, 2015-01-06 (Released:2015-01-06)

参考文献数

19

被引用文献数

3

---

This paper proposes a ranking methodology of cooking recipe by using fitness value between a recipe and onomatopoeia. This system is implemented as a function of a cooking recipe search site “Onomatoperori”. By using onomatopoeia, users can find what they want to cook from their ambiguous idea. We defined formulas for calculating fitness value between recipe and onomatopoeia by using mutual information between onomatopoeia and a word in title or description of recipes. In addition, we defined the similarity measure between onomatopeia words by mapping their words by using 15 sentimental dimensions for expressing the tastes and textures of the dishes. And we improve the ranking methodology by using the similarity among onomatopoeia words. By using these ranking methodologies we can search the cooking recipes which are related to the onomatopoeia although they do not include the onomatopeia word in the recipes.

著者

岩佐 和典 小松 孝徳

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.265-273, 2015-01-06 (Released:2015-01-06)

参考文献数

19

被引用文献数

1

---

The purpose of this study was to investigate how the naming of an object influences visually induced tactile impression and emotional valence, using tactile sense-related onomatopoeic words for evaluation of tactile impression. The present study focused on how visually induced tactile impression and valence might differ when different names of the same visual texture were presented. In addition, to investigate the relationship between visually induced tactile impression and valence, correlations among change rates of tactile impression and valence induced by the alteration of names were examined. 60 undergraduate students (mean age = 19.55, male = 3, female = 57) participated in the experiment. The results revealed that different tactile impressions and emotional responses were evoked by the same visual texture when different names were presented. Moreover, significant correlations were found between change rates of tactile impression and valence. These results show that visually induced tactile impression and emotion are influenced by top-down semantic processes.

著者

草田 裕紀 水田 孝信 早川 聡 和泉 潔

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.675-682, 2015-09-01 (Released:2015-09-01)

参考文献数

22

被引用文献数

5

---

We analyzed the impact of position-based market maker, which tries to maintain its neutral position, to the competition among stock exchanges by an artificial market simulation approach. In the previous study, we built an artificial market model and investigated for the impact of non-position-based market maker's spread to the markets' shares of trading volumes. However it had the serious problem that the non-position-based market maker is too simple to manage its own position properly and so we could not judge weather the result of previous study is correct or not. Thus in this study, we made a position-based market maker and explored the competition, in terms of taking markets' shares of trading volumes, between two artificial financial markets that have exactly the same specifications except existing a market maker, the non-position-based market maker or the position-based market maker. As a result, we found that the position-based market maker can acquire the share of trading volumes from the competitor even though its spread is bigger than bid-offer-spread of the competitor. Moreover, we revealed that position-based market maker can get a profit even in the situation that its spread or tick sizes of the stock exchanges are small. In addition to that, position-based market maker made a profit in almost all experiments which we conducted in this research by changing its spread and tick sizes of markets. At last, we confirmed that position-based market maker can manage its position properly compared to non-position-based market maker. In conclusion, the position-based market maker can not only supply liquidity to stock exchanges and contribute to acquire the share from the competitor as well as the non-position-based market maker does, but also manage its own position properly and make a profit.

著者

岡谷 英夫

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.257-264, 2015-01-06 (Released:2015-01-06)

参考文献数

17

被引用文献数

1 2

---

Sensory words such as onomatopoeia are difficult for students of Japanese because their cultures are different. How onomatopoeia are dealt with in elementary school compulsory education has been reviewed with the aim of considering how it can be applied to students of Japanese as a second language. Five Japanese textbooks that are currently in use at elementary schools for native speakers of Japanese were examined to see which onomatopoeic words appear and to what extent. A total of 6,443 onomatopoeic words were listed in these textbooks. Of the vast range of 6,443 words from the originally wide variety of words as counted from grade 1 to grade 6 from all the Japanese language textbooks, 92 were high-frequency onomatopoeic words which are proposed as the “basic onomatopoeia for beginners” as well as what kind of onomatopoeia and to what extent. These 92 high-frequency onomatopoeic words appeared 3,416 times, or 53.02% of the total 6,443 onomatopoeic words. If these 92 onomatopoeic words were studied, then over 50% of onomatopoeic words would be comprehensible to learners of Japanese. In addition, which verbs appear in conjunction with these onomatopoeic words together with their frequency are indicated.

著者

水上 雅博 Lasguido Nio 木付 英士 野村 敏男 Graham Neubig 吉野 幸一郎 Sakriani Sakti 戸田 智基 中村 哲

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-517, (Released:2015-12-15)

参考文献数

23

被引用文献数

4

---

In dialogue systems, dialogue modeling is one of the most important factors contributing to user satisfaction. Especially in example-based dialogue modeling (EBDM), effective methods for dialog example databases and selecting response utterances from examples improve dialogue quality. Conventional EBDM-based systems use example database consisting of pair of user query and system response. However, the best responses for the same user query are different depending on the user's preference. We propose an EBDM framework that predicts user satisfaction to select the best system response for the user from multiple response candidates. We define two methods for user satisfaction prediction; prediction using user query and system response pairs, and prediction using user feedback for the system response. Prediction using query/response pairs allows for evaluation of examples themselves, while prediction using user feedback can be used to adapt the system responses to user feedback. We also propose two response selection methods for example-based dialog, one static and one user adaptive, based on these satisfaction prediction methods. Experimental results showed that the proposed methods can estimate user satisfaction and adapt to user preference, improving user satisfaction score.

著者

山田 康輔 笹野 遼平 武田 浩一

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.35, no.4, pp.B-K22_1-12, 2020-07-01 (Released:2020-07-01)

参考文献数

20

---

It has been reported that a person’s remarks and behaviors reflect the person’s personality. Several recent studies have shown that textual information of user posts and user behaviors such as liking and reblogging the specific posts are useful for predicting the personality of Social Networking Service (SNS) users. However, less attention has been paid to the textual information derived from the user behaviors. In this paper, we investigate the effect of using textual information with user behaviors for personality prediction. We focus on the personality diagnosis website and make a large dataset on SNS users and their personalities by collecting users who posted the personality diagnosis on Twitter. Using this dataset, we work on personality prediction as a set of binary classification tasks. Our experiments on the personality prediction of Twitter users show that the textual information of user behaviors is more useful than the co-occurrence information of the user behaviors and the performance of prediction is strongly affected by the number of the user behaviors, which were incorporated into the prediction. We also show that user behavior information is crucial for predicting the personality of users who do not post frequently.

著者

西脇 裕作 大島 直樹 岡田 美智男

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.36, no.2, pp.B-K44_1-12, 2021-03-01 (Released:2021-03-01)

参考文献数

38

---

How can we make it possible for humans to participate in the robot’s speech to aim for a co-constructed conversation? In this paper, we investigated the effects and factors of dialogue design, focusing on “incompleteness.” We examined the people’s attitudes toward participation in multi-party conversations using “human-robot assisted story-telling” interactions. The results showed that the utterance strategy of lacking words reduced the passive participation attitude when the talker robot speak to humans directly. If we want to increase people’s participation attitude in a conversation, avoiding conveying much information and using “incompleteness” is an effective way to do so. However, the results also confirmed that the incomplete utterance was not satisfied to improve people’s co-telling attitude yet. The robots in this study were unable to accept the variety of ways in which people speak. To achieve the co-constructed conversation, discussed how robots could install a variety of actions based on other multi-party conversation studies. Therefore, we also investigated the limitation of multi-party participation and the characteristics of human speeches. For people and systems to have a co-constructed conversation rather than as information transfer, we believe that the design of dialogue needs to change. For this reason, we reported one of the effects of “incompleteness” conversation design here.

著者

廣中 詩織 吉田 光男 梅村 恭司

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.35, no.1, pp.E-J71_1-10, 2020-01-01 (Released:2020-01-01)

参考文献数

26

---

Users’ attributes, such as home location, are necessary for various applications, such as news recommendations and event detections. However, most real user attributes (e.g., home location) are not open to the public. Therefore, their attributes are estimated by relationships between users. A social graph constructed from relationships between users can help estimate home locations, but it is difficult to collect many relationships, such as followers’ relationships. We focus on users whose home locations are difficult to estimate, so that we can select users whose locations can be accurately estimated before collecting relationships. In this paper, we use their profiles which can be collected before collecting relationships. Then, we analyze difficult users with their profiles. As a result, we found that users whose home locations incorrectly estimated had a longer duration since the date their account was created, longer name, and longer description. In addition, the results indicated that the users whose home locations were incorrectly estimated differed from those whose home locations could not be estimated.

著者

大澤 昇平 松尾 豊

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.5, pp.469-482, 2014-09-01 (Released:2014-08-15)

参考文献数

22

被引用文献数

1

---

In social networking service (SNS), popularity of an entity (e.g., person, company and place) roles an important criterion for people and organizations, and several studies pose to predict the popularity. Although recent papers which addressing the problem of predicting popularity use the attributes of entity itself, typically, the popularity of entities depends on the attributes of other semantically related entities. Hence, we take an approach exploiting the background semantic structure of the entities. Usually, many factors affect a person's popularity: the occupation, the parents, the birthplace, etc. All affect popularity. Predicting the popularity with the semantic structure is almost equivalent to solving the question: What type of relation most affects user preferences for an entity on a social medium? Our proposed method for popularity prediction is presented herein for predicting popularity, on a social medium of a given entity as a function of information of semantically related entities using DBpedia as a data source. DBpedia is a large semantic network produced by the semantic web community. The method has two techniques: (1) integrating accounts on SNS and DBpedia and (2) feature generation based on relations among entities. This is the first paper to propose an analysis method for SNS using semantic network.

著者

高野 雅典 和田 計也 福田 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

2015

被引用文献数

3

---

Cooperative behaviors are common in humans and are fundamental to our society. Theoretical and experimental studies have modeled environments in which the behaviors of humans, or agents, have been restricted to analyze their social behavior. However, it is important that such studies are generalized to less restrictive environments to understand human society. Social network games (SNGs) provide a particularly powerful tool for the quantitative study of human behavior. In SNGs, numerous players can behave more freely than in the environments used in previous studies; moreover, their relationships include apparent conflicts of interest and every action can be recorded. We focused on reciprocal altruism, one of the mechanisms that generate cooperative behavior. This study aims to investigate cooperative behavior based on reciprocal altruism in a less restrictive environment. For this purpose, we analyzed the social behavior underlying such cooperative behavior in an SNG. We focused on a game scenario in which the relationship between the players was similar to that in the Leader game. We defined cooperative behaviors by constructing a payoff matrix in the scenario. The results showed that players maintained cooperative behavior based on reciprocal altruism, and cooperators received more advantages than noncooperators. We found that players constructed reciprocal relationships based on two types of interactions, cooperative behavior and unproductive communication.

著者

楊 碩 橋本 敬 李 冠宏 李 暁燕

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.331-339, 2015-01-06 (Released:2015-01-06)

参考文献数

20

---

Japanese onomatopoeia is an important element to express feelings and experiences lively. It is very difficult for Japanese learners to acquire onomatopoeia, especially, its nuance. In this paper, based on traditional L2 learning theories, we propose a new learning method to improve the efficiency of learning Japanese onomatopoeias' nuance - both explicit and implicit - for non-native speakers. The method for learning implicit nuance of onomatopoeia consists of three elements. First is studying the formal rules representing the explicit nuances of onomatopoeic words. Second is creating new onomatopoeic words by learners to utilize those formal rules. The last is giving feedback of relevance of the onomatopoeias created. We then show a learning system implementing the proposed method. In addition, to verify the effectiveness of the proposed method and the learning system, we conducted an experiment involving two groups of subjects. While the experiment group covers all the three elements of the proposed method, the control group involves no creation process, which is supposed to be a core element of our proposed method, instead, does an assessment process in which the participants assess the appropriateness of onomatopoeic words presented. Both groups were required to take two tests, before and after going through the learning process. The learning effect is defined as the difference between the scores gained from pre-learning test and post-learning test. The result confirms that the proposed method has significant effect in learning onomatopoeia for non-native speakers. Moreover, the comparison against the control group shows that the creation process is the key to bring the learning effect.

著者

岡谷 英夫

出版者

The Japanese Society for Artificial Intelligence

雑誌

人工知能学会論文誌 (ISSN:13460714)

巻号頁・発行日

vol.30, no.1, pp.257-264, 2015

被引用文献数

2

---

Sensory words such as onomatopoeia are difficult for students of Japanese because their cultures are different. How onomatopoeia are dealt with in elementary school compulsory education has been reviewed with the aim of considering how it can be applied to students of Japanese as a second language. Five Japanese textbooks that are currently in use at elementary schools for native speakers of Japanese were examined to see which onomatopoeic words appear and to what extent. A total of 6,443 onomatopoeic words were listed in these textbooks. Of the vast range of 6,443 words from the originally wide variety of words as counted from grade 1 to grade 6 from all the Japanese language textbooks, 92 were high-frequency onomatopoeic words which are proposed as the "basic onomatopoeia for beginners" as well as what kind of onomatopoeia and to what extent. These 92 high-frequency onomatopoeic words appeared 3,416 times, or 53.02% of the total 6,443 onomatopoeic words. If these 92 onomatopoeic words were studied, then over 50% of onomatopoeic words would be comprehensible to learners of Japanese. In addition, which verbs appear in conjunction with these onomatopoeic words together with their frequency are indicated.

著者

玉川 奨 桜井 慎弥 手島 拓也 森田 武史 和泉 憲明 山口 高平

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.25, no.5, pp.623-636, 2010 (Released:2010-08-17)

参考文献数

18

被引用文献数

4 3

---

Here is discussed how to learn a large scale of ontology from Japanese Wikipedia. The learned ontology includes the following properties: rdfs:subClassOf (IS-A relationship), rdf:type (class-instance relationship), owl:Object/DatatypeProperty (Infobox triple), rdfs:domain (property domain), and skos:altLabel (synonym). Experimental case studies show us that the learned Japanese Wikipedia Ontology goes better than already existing general linguistic ontologies, such as EDR and Japanese WordNet, from the points of building costs and structure information richness.

著者

岡田 将吾 松儀 良広 中野 有紀子 林 佑樹 黄 宏軒 高瀬 裕 新田 克己

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.AI30-E_1-12, 2016-11-01 (Released:2016-11-02)

参考文献数

22

被引用文献数

6

---

This paper focuses on developing a model for estimating communication skills of each participant in a group from multimodal (verbal and nonverbal) features. For this purpose, we use a multimodal group meeting corpus including audio signal data and head motion sensor data of participants observed in 30 group meeting sessions. The corpus also includes the communication skills of each participant, which is assessed by 21 external observers with the experience of human resource management. We extracted various kinds of features such as spoken utterances, acoustic features, speaking turns and the amount of head motion to estimate the communication skills. First, we created a regression model to infer the level of communication skills from these features using support vector regression to evaluate the estimation accuracy of the communication skills. Second, we created a binary (high or low) classification model using support vector machine. Experiment results show that the multimodal model achieved 0.62 in R2 as the regression accuracy of overall skill, and also achieved 0.93 as the classification accuracy. This paper reports effective features in predicting the level of communication skill and shows that these features are also useful in characterizing the difference between the participants who have high level communication skills and those who do not.

著者

松尾 豊 友部 博教 橋田 浩一 中島 秀之 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.20, no.1, pp.46-56, 2005 (Released:2005-01-05)

参考文献数

20

被引用文献数

13 28

---

Social relation plays an important role in a real community.

著者

笹原 和俊

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

pp.B-MDF02, (Released:2016-01-28)

参考文献数

16

---

The socialization of the Web changes the ways we behave both online and offline, leading to a novel emergent phenomenon called ``collective attention'' in which people's attention is suddenly concentrated on a particular real-life event. Visualizing collective attention is fundamental to understand human behavior in the digital age. Here we propose ``association networks'' to visualize usage-based, term-association patterns in a large dataset of tweets (short text messages) during collective attention events. First, we train the word2vec model to obtain vector representations of terms (words) based on semantic similarities, and then construct association networks: given some terms as seeds, the associated terms are linked with each other using the trained word2vec model, and considering the resulting terms as new seeds, the same procedure is repeated. Using two sets of Twitter data---the 2011 Japan earthquake and the 2011 FIFA Women's World Cup---we demonstrate how association networks visualize collective attention on these events. Provided the Japan earthquake dataset, the association networks that emerged from the most frequently used terms, such as earthquake and tsunami, exhibit distinct network structure related to people's attention during the earthquake, whereas one that emerged from emotion-related terms, such as great and terrible, shows a large connected cluster of negative terms and small clusters of positive terms. Furthermore, we compare association networks in different datasets, using the same seed terms. These results indicate the proposed method to be a useful tool for visualizing the implicit nature of collective attention that is otherwise invisible.