笹嶋 宗彦

<sec> <title>Background</title> Chatbots are artificial intelligence–driven programs that interact with people. The applications of this technology include the collection and delivery of information, generation of and responding to inquiries, collection of end user feedback, and the delivery of personalized health and medical information to patients through cellphone- and web-based platforms. However, no chatbots have been developed for patients with lung cancer and their caregivers. </sec> <sec> <title>Objective</title> This study aimed to develop and evaluate the early feasibility of a chatbot designed to improve the knowledge of symptom management among patients with lung cancer in Japan and their caregivers. </sec> <sec> <title>Methods</title> We conducted a sequential mixed methods study that included a web-based anonymized questionnaire survey administered to physicians and paramedics from June to July 2019 (phase 1). Two physicians conducted a content analysis of the questionnaire to curate frequently asked questions (FAQs; phase 2). Based on these FAQs, we developed and integrated a chatbot into a social network service (phase 3). The physicians and paramedics involved in phase I then tested this chatbot (α test; phase 4). Thereafter, patients with lung cancer and their caregivers tested this chatbot (β test; phase 5). </sec> <sec> <title>Results</title> We obtained 246 questions from 15 health care providers in phase 1. We curated 91 FAQs and their corresponding responses in phase 2. In total, 11 patients and 1 caregiver participated in the β test in phase 5. The participants were asked 60 questions, 8 (13%) of which did not match the appropriate categories. After the β test, 7 (64%) participants responded to the postexperimental questionnaire. The mean satisfaction score was 2.7 (SD 0.5) points out of 5. </sec> <sec> <title>Conclusions</title> Medical staff providing care to patients with lung cancer can use the categories specified in this chatbot to educate patients on how they can manage their symptoms. Further studies are required to improve chatbots in terms of interaction with patients. </sec>

Nurses’ daily practices are performed according to nursing guidelines of hospitals. Novice nurses learn by consulting the bulky files of the guidelines for their practices. The nurses ’actions in practice are usually transmitted from experienced nurses to the novices through the daily practices. It is difficult to represent complex and intertwined actions with their goals by the paper guidelines. The goal of the actions is often implicit knowledge accumulated through the nurses ’daily practices. To cope with this problem, we have previously proposed a goal realization model called“ CHARM ”(an abbreviation for“ Convincing Human Action Rationalized Model ”). Unlike in the printed guidelines, CHARM can represent the goal of actions. We have described nursing actions and their goals as CHARM trees constructed from real nursing guidelines used in a hospital through consultation with experienced nurses. On the other hand, the novices need to learn sequence of the actions. Therefore we developed a tablet computer, which is called CHARM Pad, for browsing CHARM trees with two modes corresponding to the both viewpoints. The novices are able to learn nursing actions with their goals by both goal-oriented viewpoint and sequence-oriented viewpoint. The CHARM Pad has been used for training of the novice nurses at ICU unit. After the training, we received positive comments from the chief nurse in the ICU unit and the chief nurse in the Nursing Department. Questionnaires completed by the novice nurses also showed that the two modes of the CHARM Pad helped them to understand the goals and the sequence of the actions.

Procedural knowledge about human actions is usually documented in manuals in the form of natural language. In the nursing domain, the standard procedures of nursing actions are also described in nursing manuals, which are used for training of novice nurses. The issue that we focus on in this research is that the rationale for the actions, i.e., "why the actions should be done as specified in the procedure", tends to be implicit in these manuals. Some novice nurses may not be able to perform suitable actions adaptively in response to an irregular situation because they do not understand such rationale. To cope with this problem, we have previously proposed a goal realization model called "CHARM" (an abbreviation for "Convincing Human Action Rationalized Model"). We have described nursing actions and their goals as CHARM trees constructed from real nursing manuals used in a hospital through consultation with experienced nurses. We have also developed a tablet-style tool for browsing CHARM trees, which has been successfully deployed in a training course for novice nurses. The aim of the work described here was to develop a method for systematically describing pieces of action knowledge organized in a knowledge base (KB) for future reuse in describing CHARM trees. By reusing the pieces of knowledge in the KB, the knowledge authors can easily build different CHARM trees. The key element for systematically describing pieces of action knowledge is a "way of action achievement", which represents goal realization knowledge, i.e., "how to achieve an action as a state change". Its description consists of a goal-action as a goal state to be achieved, a sequence of sub-actions that can achieve the goal-state, and background knowledge such as physical principles. In this paper, we demonstrate how to build such a reusable and systematic KB for nursing actions. We extracted the ways of action achievement from the real CHARM trees mentioned above. The ways of action achievement for a specific action were systematically organized in a hierarchy of categories. The contributions of this work to the knowledge management community include (1) a knowledge modeling framework of human actions for knowledge sharing and training, (2) a well-founded method for systematically describing ways of action achievement, and (3) practical demonstrations of this framework and method using real knowledge in the nursing domain.

The value of information accumulated on the Web is enhanced when it is provided to the user who faces a problematic situation that can be solved by the information. The authors have investigated a taskoriented menu that enables users to search for mobile Internet services not by category but by situation. Construction of the task-oriented menu is based on a user modeling method that supports descriptions of user activities, such as task execution and defeating obstacles encountered during the task, which in turn represents the users' situations and/or needs for certain information. They built task models of the mobile users that cover about 97% of the assumed situations of mobile Internet services. Then they reorganized "contexts" in the model and designed a menu hierarchy from the viewpoint of the task. The authors have linked the designed menu to the set of mobile Internet service sites included in the i-mode service operated by NTT docomo, consisting of 5016 services. Among them, 4817 services are properly connected to the menu. This chapter introduces a framework for a real scale task-oriented menu system for mobile service navigation with its relations to the SNS applications as knowledge resources. © 2013, IGI Global.

Japan is no exception among developed countries facing healthcare system problems due to aging and low birthrate as the number of patient increases and health care worker numbers shrink. The introduction of high tech medicine has increased the amount of knowledge to be learned by novice nurses. Although vast amounts of implicit knowledge have accumulated among nursing practitioners, this knowledge needs to be communicated when hospitals train younger generations of nurses. In this paper, the authors have proposed an activity model called CHARM (the Convincing Human Action Rationalized Model) and CHARM Pad, tablet PC with browsing software for CHARM models. CHARM explicates multidimensional purpose-oriented procedure relations often existing as implicit knowledge. For this reason, CHARM supports the training and education of novice nurses. We developed CHARM models according to nursing guidelines of hospitals and applied them to training of the nurses in two hospitals. CHARM and CHARM Pad are being evaluated at these hospitals and positive responses are coming from nurses. © 2013 Springer-Verlag Berlin Heidelberg.

Service research is a kind of multidisciplinary research including management, marketing, economics, engineering and so on. "Services" are one of the fundamental concepts in all disciplines that compose service research. Many researchers have proposed definitions of services to date. The adequacy of these existing definitions, however, has never been evaluated. In this paper, we evaluate the adequacy of 13 existing definitions of the concept of services based on ontology engineering. Firstly, we analyze the conceptual structures of these definitions of services by clarifying their base concepts and attributes. The attributes are related to functions as shown in our previous research where we showed the notion that a function is a fundamental concept of services. Next, we discuss the evaluation criteria and evaluate the existing definitions based on these criteria. The result of the evaluation shows that the existing definitions are not adequate according to the criteria. Moreover, we evaluate our definition of services proposed in our previous research in the same way, and confirm its adequacy.

看護師が看護現場での実践力を高めるためには，マニュアルに記述された看護手順を盲目的に学ぶだけでなく，手順の目的や状況に応じた代替手段， 回避すべきリスクなど看護知識を体系的に学ぶことも必要である．本発表では，看護の実践力の源となるべく筆者らが開発を進めているツール CHARM Padを，実際の病院への導入および試用事例と合わせて紹介する．

Although many definitions of services have been proposed in Service Science and Service Engineering, essentialities of the notion of 'service' remain unclear. Especially, some existing definitions of service are similar to the definition of function of artifacts, and there is no clear distinction between them. Thus, aiming at an ontological conceptualization of service, we have made an ontological investigation into the distinction between service and artifact function. In this article, we reveal essential properties of service and propose a model and a definition of service. Firstly, we extract 42 properties of service from 15 articles in different disciplines in order to find out fundamental concepts of service. Then we show that the notion of function shares the extracted foundational concepts of service and thus point out the necessity of the distinction between them. Secondly, we propose a multi-layered model of services, which is based on the conceptualization of goal-oriented effects at the base-level and at the upper-level. Thirdly, based on the model, we clarify essential properties of service which can distinguish artifact function. The conceptualization of upper-effects (upper-service) enables us to show that upper-services include various effects such as sales and manufacturing. Lastly, we propose a definition of the notion of service based on the essential properties and show its validity using some examples.

This paper proposes an ontological definition of services. Such a definition is one of the fundamentals of service research. The understanding of essentialities of the notion of services as its definition, which differentiates services from the other notions, contributes proper modeling and conceptualization of services in services design and knowledge management of services. The existing definitions and characteristics of services, however, cannot differentiate services from other concepts: especially function of product. In this paper, we propose a new definition of services based on ontological consideration. Our definition can differentiate services from product functions. Firstly, we discuss the problem of the existing definitions and characteristics about the distinction. Secondly, we explain our definition and the essential characteristics of services using an ontological model of services. Lastly, we demonstrate its applicability of our definition using some examples and compare it with the existing definitions.

In order to facilitate sharing of functional models, some functional taxonomies each of which provides a set of verbs for representing generic functions (called functional terms here) have been developed. Their examples include some sets of generally valid functions in the book written by Pahl and Beitz, Functional Basis (FB) developed by Hirtz et al. and FOCUS/Tx developed by the authors of this paper. The issue addressed in this paper is the implicitness of the criteria of classification of functional terms in those taxonomies and thus unclearness of their definitions. This paper proposes an ontology of logical criteria for classification of functional terms (called FOCUS/View). Using the classes defined in FOCUS/View, the classification criteria of a functional taxonomy can be explicitly represented. These classes have been conceptualized based on deep investigation on FB and FOCUS/Tx. The benefits of the proposed FOCUS/View ontology include: (1) users of a taxonomy can easily understand differences of similar terms and select an appropriate term out of them, (2) a developer Of a taxonomy can check its logical classification structure and then improve its logical clearness, and (3) we can compare different taxonomies and establish more reliable mappings between their terms for interoperability of functional models. In this paper, as a demonstration of the benefit (1), the classification criteria of FB, FOCUS/Tx, Krumhauer's and Roth's generally valid functions are explicitly presented using FOCUS/View. As a demonstration of the benefit (2), some logically problematic classification structures of FB and the Krumhauer's functions are discussed and modified for the logical clearness. For the benefit (3), this paper demonstrates the mappings between FB and FOCUS/Tx and a semantic interoperable document search system based on these mappings.

New in-vehicle information provision services have started according to the progress of data communication infrastructure surrounding vehicles, where the vehicles can connect to communication network outside. In such information services, a large amount of various data related to vehicles and/or drivers will be stored to the datacenter over the network in near future. Data and functions which are currently equipped with in-vehicle devices will be shifted to the datacenter, and these data and functions will be delivered to each of vehicles from the data center in need. In this paper, we propose a method of structuring such data based on ontological engineering approach for the purpose of providing drivers with useful information utilizing a large amount of communication data effectively. That is, we introduce a mechanism of providing drivers with information effectively, using the data structure which refers to ontology as a data schema. Also we describe about the information space with the contents of information to provide and the driver model with ontology. © 2010 IEEE.

Value of information accumulated on the Web should be enhanced if it is provided to the user who just faces to a problematic situation which can be solved by the information. The authors have been investigating a task-oriented menu, which enables users to search for mobile internet services not by category but by situation of the users. Construction of the task-oriented menu is based on a user modeling method which supports descriptions of user activities, such as task execution and defeating obstacles encountered during the task, which in turn represents users&apos; situations and/or needs for certain information. We have built task models of the mobile users which covered about 97% of the assumed situations of mobile internet services. Then we reorganized "contexts" in the model and designed a menu hierarchy from the view point of the task. We have linked the designed menu to the set of actual mobile interne service sites included in the i-mode service operated by NTT docomo, consists of 5016 services. Among them, 4817 services are properly connected to the menu. This paper introduces a framework for real scale task-oriented menu system for mobile service navigation with its relations to the SNS applications as knowledge resources.

Growth in the mobile internet services industry has seen a marked increase in the number of mobile internet services provided, and this has made proper structuring and organization of the services difficult. Present methods of service provision have proven insufficient to guide users efficiently to the services they need. To solve this problem, a task-oriented menu, which enables users to search for services by what they want to do and not by category, has been proposed. Construction of the task-oriented menu is based on a task ontology modeling method which supports descriptions of user activities, such as task execution and defeating obstacles encountered during the task. This paper discusses a task ontology-based modeling method which supports descriptions of users' activities and related knowledge, such as how to solve problems that the users encounter and how to prevent or solve them on the spot. Models described by our method contribute to designing, testing, and improving mobile internet services.

This paper discusses ontology mapping between two taxonomies of functions of artifacts for the engineering knowledge management. The mapping is of two ways and has been manually established with deep semantic analysis based on a reference ontology of function for bridging the ontological gaps between the taxonomies. We report on the successful results thanks to such deep analysis not at the lexical level but at the ontological level. Using the mapping knowledge, we developed a semantic search system which can provide engineers with interoperable access to technical documents by searching for functional metadata based on either of functional taxonomies. © 2008 Springer Berlin Heidelberg.

Ten years have passed since the concept of the semantic web was proposed by Tim Berners-Lee. For these years, basic technologies for them such as RDF(S) and OWL were published. As a result, many systems using semantic technologies have been developed. Some of them are not prototype systems for researches but real systems for practical use. The authors analyzed semantic web applications published in the semantic web conferences (ISWC, ESWC, ASWC) and classified them based on ontological engineering. This paper is a review of application papers published in Semantic Web conferences. We discuss a trend and the future view of them using the results.