Research on Informatics for Learning and Education
Akihiro Kashihara*
Abstract The key to developing learning support using technology is understanding learning from an informatics point of view.
Research on informatics for learning and education applies learning theories or views to model learning and develop various methods and systems for learning support. The Japanese Society for Information and Systems in Education (JSiSE) has demonstrated leader-ship in this research field in Japan. JSiSE particularly encourages theoretical development, as well as practical use of models, meth-ods, and systems, which involves circular interaction between them. In this paper, we introduce the history, research aim and approach, and academic activities of JSiSE. The paper also highlights the research being conducted by JSiSE on theoretical modeling of learning and development of methods and systems, and its transition owing to the influence of learning theory and learning media.
Keywords modeling of learning, theoretical development, practical use, design approach, JSiSE
1. Introduction
Learning is certainly difficult for people. Numerous studies on learning and education have addressed the essential issue of how to support learning using technol-ogy[1]. The key to technological support for learning is to understand what learning is and how people learn from an informatics point of view[2], [3]. Most studies have accordingly adopted learning theories developed in related fields such as cognitive science to model learn-ing and develop methods and systems for learnlearn-ing sup-port[4]. Theoretical modeling and development involve exploring ways to represent, analyze, generate, transmit, exchange, and operate information related to learning and its support. These also require synthesizing infor-mation into methods and systems. Such informatics approach to information and systems for learning and education using technology has been investigated[1], [3].
On the other hand, technological progress in media (learning media) for learning such as PC, mobile phone, the Internet, tablet PC, VR, and robot, has led to diver-sity in learning aspects and styles. However, theoreti-cally understanding these diversified learning aspects or styles is difficult. Another approach is to design a learn-ing model that reflects how people should learn uslearn-ing learning media[3], [4]. This learning model allows us to understand learning from the viewpoint of not how real learners learn but how to intentionally induce them to learn using technology. Since learning is increasingly becoming diversified, the use of the model design
approach is becoming feasible and important in infor-matics for learning and education.
In Japan, the Japanese Society for Information and Systems in Education[5] (JSiSE) has initiated and sup-ported research in the field of informatics for learning and education since 1974. To resolve the essential research question of how to support learning, JSiSE has established a research community, and provided researchers opportunities for various academic services related to journals, annual conference, SIG meetings, industry–academia collaboration, support for younger researchers, etc. In addition, JSiSE encourages theoreti-cal development as well as practitheoreti-cal use of models, methods, and systems. It particularly promotes circular interaction between them in which theoretical develop-ment is evaluated in practice, the results obtained in practice are expected to lead to new questions, and questions that arise in practice are addressed in theoreti-cal development. Such circular interaction is essential for fostering research in the field of informatics for learning and education.
In the following sections of this paper, I introduce the JSiSE, including history, research aim and approach, and academic activities. The paper also describes the research conducted by JSiSE on the theoretical modeling of learning and development of methods and systems, and its transition as a result of the influence of learning theory and learning media[3]. Finally, the paper elabo-rates on future research directions.
* Graduate School of Informatics and Engineering, The University of Electro-Communications, Japan
President of the Japanese Society for Information Systems in Education
2. JSiSE
2.1 History
JSiSE was founded in 1974 as an academic society for researchers in computers in education before the wide-spread use of PCs. It started with the name Japan Society for CAI (Computer-Assisted Instruction). It has a long-standing history in Japan in terms of research on learning and education using technology. Initially, the main research interest of JSiSE was the use of comput-ers as learning media to provide instructions to learncomput-ers. The learning media used have been gradually extended to diverse information and communication technologies (ICTs) including PCs, the Internet, and mobile phones. In 1995, the name of the society was changed to Japanese Society for Information and Systems in Education. Subsequently, it was approved as a general incorporated association in 2014. This author became a student member of this society in 1988, and has been an active member for more than 30 years.
JSiSE currently has approximately 1,200 members. Although its members are not many, a considerable pro-portion of its members actively participate in major international conferences related to learning and educa-tional technologies such as AIED, ICCE, IEEE-ICALT, and CSCL. Some key members make remarkable contri-butions to international societies such as APSCE[6] and IAIED[7].
JSiSE consists of an executive committee (EC) and several permanent committees with individual missions including journal publishing, annual conference hosting, special interest group (SIG) building, and mentoring of young researchers. Each EC member is assigned to one of the permanent committees to carry out a specific mis-sion. JSiSE currently has seven branches nationwide, namely Hokkaido, Kanto, Hokushin-etsu, Tokai, Kansai, Chugoku, and Kyushu. Each branch performs activities for both researchers and students in its region.
2.2 Aim and Approach
JSiSE has initiated research on learning and education, particularly using ICTs. The focus of the research is not merely to apply existing ICTs or educational applica-tions to educational systems or learning environments without any deliberate consideration of how learners learn, since instructive support for learning requires
understanding about learning. In terms of technological learning support, the research field aims to follow learn-ing theories developed in related fields, such as psychol-ogy, cognitive science, and social science, to elucidate information processing on what learning is and how it should be supported[3]. These correspond to developing learning models, methods, and systems for learning and education. JSiSE has encouraged such theoretical devel-opments.
In addition, JSiSE also promotes practical use of models, methods, and systems developed. Fig. 1 shows the flow of theoretical development and practice. In this flow, models, methods, and systems are tested and eval-uated in practice. The results obtained from practice are expected to contribute to refining the theoretical devel-opment and give rise to new research questions. These questions are in turn expected to lead to new theoretical development. This interaction between theoretical development and practice can become a driving force for progress in the research field.
2.3 Activities
JSiSE offers various activities to its members to pro-mote theoretical development and practical use of mod-els, methods, and systems for learning and education. First, it has been publishing Japanese transactions and an English journal known as The Journal of
Information and Systems in Education. The
transac-tions began publication in 1980 as the Journal of Japan
Society for CAI. The name was changed to its current
one in 1995 and has so far been increased to volume 37. Papers can be open-accessed on J-STAGE one year after they are published[8]. The English journal has also been Figure 1. Flow of theoretical development and practical use of
published yearly since 2002 and has increased to vol-ume 19. It is also an open access journal on J-STAGE[9].
Both the transactions and English journal include four paper categories: regular, practical, short note, and report on practice. Regular and practical papers corre-spond to what is known as original article. Short note and report on practice papers aim to promptly publish theoretical development and practice. Practical and report on practice papers are particularly expected to present some research questions and their results, allow-ing researchers to explore new theoretical develop-ments. One or two papers are selected for the best paper award from those published as regular and short note in even years and from those published as practical and report on practice in odd years. JSiSE enables research-ers to conduct studies not only on theoretical develop-ment but also practice via these paper categories, and it further facilitates circular interaction between them.
Besides, the editorial committee plans and pub-lishes a special issue of the transactions every year in cooperation with the committee in charge of key research topics JSiSE aims to address in the near future. The special issues in the last three years are as follow:
- 2020 special issue on a New Paradigm for Education and Learning Management using Artificial Intelligence and IoT Techniques (Vol. 37, No. 2),
- 2019 special issue on Learning Support Environments Utilizing Media/Devices Based on New Technologies (Vol. 36, No. 2), and
- 2018 special issue on Educational System for Safe and Secure Society, Practical/Support Systems for Programming and Information Technology Education (Vol. 35, No. 2).
The editorial committee of the transactions has just initiated an interesting plan to devote a small number of pages to enumerate the research questions that the papers published in each issue addressed. The published papers tend to focus on how to address the research questions; however, it is more important to clarify the questions themselves. This approach is expected to dis-tribute research questions among JSiSE researchers to compare and distinguish their individual works. It is also expected to allow newcomers to start up their work in JSiSE. Such question distribution is expected to cre-ate a vantage point for overviewing the research field of
JSiSE in the future.
In terms of the English journal, it will be integrated into an international journal known as Information and
Technology in Education and Learning (ITEL) from
2021. ITEL is published as a joint journal with the Japanese Society for Educational Technology (JSET). JSiSE hopes that this collaboration with JSET will strengthen the practical aspects of research in JSiSE.
JSiSE holds SIG meetings and annual conferences. The research interests of SIGs are quite diverse, from theoretical development to practice, including the design of learning environments, model for practice, learning analytics, advanced educational use for medical care, nursing, and welfare, assessment, AI in education (AIED), advanced learning technology, computer sci-ence/programming education, and e-learning practice. Every year, JSiSE holds six SIG meetings and another meeting focusing on the special issue published in the transactions, which publish a total of approximately 150 papers. JSiSE selects 3% of these papers for the best paper awards.
The annual conference is held for three days under a theme representing a hot topic in the research field. The conference themes in the last three years are as fol-lows:
- 45th Annual Conference Theme (2020): Present and Future of Educational Systems for Sustainable Learning,
- 44th Annual Conference Theme (2019): Towards Resilient Learning by means of Technology for Sharing Knowledge and Intelligence, and
- 43rd Annual Conference Theme (2018): Collaboration between Industry and Academia Creates Future Platform for Education and Learning.
The annual conference creates an opportunity for members to meet face to face once a year, and more than 350 members gather to discuss and build academic relationships. It consists of a keynote speech, regular sessions, interactive poster sessions, SIG sessions, spe-cial session for students, enterprise session, and sympo-sium presented by the JSiSE branch conference host.
In addition to these academic activities, JSiSE pro-motes cooperation between the industry and academia. The key activity is a competition on learning technology known as Learning Innovation Grand Prix (LIGP)[10],
which has been held since 2016. LIGP is a unique com-petition in Japan in which researchers and practitioners in enterprises evaluate and rank academic results pro-duced by academic research groups. The evaluation cri-teria are: (1) industrial novelty and creativity and (2) social values including impact on education, scale of idea, and practical effects expected on education. Thereafter, a group is selected as the grand prix, and two groups as the semi grand prix. For academic research-ers, the comments and suggestions from LIGP can bring about beneficial findings and awareness, which are dif-ferent from those from an academic point of view, to help refine their work. The grand prix winners since 2016 are as follows:
- LIGP 2019 Grand Prix: Evaluating Web-based Investigative Learning without Correct Answer, presented by the Kashihara Lab, The University of Electro-Communications,
- LIGP 2018 Grand Prix: An Attempt to Diagnosing Mental States from Biometric Information with Machine Learning in Real-time, presented by the Matsui Lab, Waseda University,
- LIGP 2017 Grand Prix: Recoco: An Application of Voice Memo for Reflection, presented by the Recoco Team, Tsukuba University, and
- LIGP 2016 Grand Prix: TED SYSTEM: A Training System for Tsunami Drill, presented by Team Mitsuhara, Tokushima University.
3.
Research on Informatics for
Learning and Education
To support learning, it is indispensable to understand learning. In particular, systematic support with ICTs requires viewing learning as information processing. This introduces an essential issue of how to model learning from an informatics point of view. Many researchers have developed learning models and they have consequently used these models to develop meth-ods and systems for learning. Such theoretical develop-ment has been influenced by the use of learning theory and learning media.
In the following subsections, I will first present methods for modeling of learning, which have been employed so far in research on informatics for learning and education. I then present the transition of research
on theoretical development owing to the influence of learning theories with an example in the AIED field. In addition, the influence of the learning media on the tran-sition of research is described.
3.1 Modeling of Learning
There are two approaches to modeling of learning: ana-lytical approach and design approach. Anaana-lytical approach focuses on authenticity in terms of how learn-ers learn, which has been investigated in related studies on cognitive science including learning science. It was considered to be the main approach at the early stages of research on informatics for learning and education. Under this approach, researchers use learning theories developed in related fields to model the cognitive pro-cess in learning as authentic as possible[11], which involves exploring the ways to represent, analyze, gen-erate, transmit, exchange, and operate information related to learning. This can result in scientific findings in learning, which are important in informatics. However, authentic modeling requires accurate evalua-tion, which is not easily realized. In addievalua-tion, technolog-ical progress in learning media has led to diversity in learning aspects and styles. Accordingly, it is more com-plicated to model learning authentically. Therefore, modeling diversified learning would become restrictive.
The design approach to modeling of learning, on the other hand, places emphasis on how learners should learn. Simon noted that designing a cognitive system offers an instructive way to understand cognition[12]. This idea can be applied to modeling of learning. The model design allows researchers to understand learning from the viewpoint of how learners should learn using learning media. In deciding the objective, it is indis-pensable to grasp the aspects and styles of learning for which learning media could be used effectively[3], [4]. In this approach, the expected results are not authentic models but are learning benefits obtained from learning as modeled. The significance of designed models is dependent on whether learners derive some benefits by learning as modeled.
To allow learners learn as modeled, the importance of external representation for a learning model should be emphasized, from which one can externally operate information representing the knowledge used in prob-lem-solving or thinking related to learning, and describe states, processes, and results of their learning[4]. We
have also developed methods and systems with such external representations, which allow learners to exter-nalize their learning process as modeled[13], [14].
For instance, we have designed a model of investi-gative learning with a question on the Web[14]. It includes three cyclic phases, which are search for Web resources related to the question, navigation and knowl-edge construction about the question, and question expansion. Although such Web-based investigative learning is generally supported by a Web browser, we have developed a cognitive tool involving scaffolding for these phases, which is implemented as an ad-on for Firefox, as shown in Fig. 2. This tool allows learners to search for and navigate Web resources/pages, construct their knowledge about the initial question, and expand it into sub-questions using a question tree. It also allows them to cyclically investigate the sub-questions within the three phases. In this way, the cognitive tool allows learners to conduct the Web-based investigative learning as modeled.
These analytical and design approaches to model-ing of learnmodel-ing have so far been investigated by JSiSE.
Although researchers in JSiSE have also been develop-ing new learndevelop-ing media to explore models of learndevelop-ing for which they could be suitable, the learning aspects and styles are increasingly diversified owing to techno-logical progress. It is difficult to preserve authenticity in modeling such diversified learning. The design approach, however, becomes more promising and feasi-Figure 2. Cognitive tool for model-based investigative learning on the Web
Figure 3. Transition of existing learning theories/views (modified from [3])
ble as it allows researchers to intentionally decide how learners should learn using diverse learning media[3].
3.2 Influence of Learning Theories
Learning theory or view of learning has significant influ-ence on how learning is modeled and supported[12], [15]. Fig. 3 shows the transition of the main existing learning theories or views in research on learning and education using technology[3].
Before the 1970s, behaviorism was the only exist-ing learnexist-ing theory in the research field of learnexist-ing and education. In this view, learning is considered to a behaviorist model. Early researchers used this model to develop CAI systems in which they mainly explored how instructional information including problems, explanation, and hints, can be presented according to learners’ responses. They aimed to implement an indi-vidualized environment for instruction using computers. At present, theoretical development through this view of learning has been replaced with e-Learning systems.
Since the 1970s, there are diverse existing learning theories or views in the research field of informatics for learning and education. There has also been a paradigm shift on how systems are developed, giving rise to vari-ous types of learning environments and educational sys-tems.
For example, the learning view of AIED research changed from behaviorism to cognitivism in the 1970s. This view change is extremely important for this research field to start exploring intelligent learning sup-port[1], [15]. In the cognitivist view, learning is considered to be a cognitive model. This emphasizes the impor-tance of describing the cognitive process in problem-solving and thinking related to learning. Researchers in AIED have accordingly explored methods for identify-ing and modelidentify-ing the learners’ cognitive process. In par-ticular, attempts have been made to represent bugs in knowledge and errors in problem-solving process[16]. Moreover, methods for adapting instructional informa-tion presented to individual learners considering their bugs and errors have been developed. Systems including such learner modeling and adaptive support are classi-fied into ITS (Intelligent Tutoring System)[15]. ITS, in addition to CAI, focuses on communicating the knowl-edge necessary for learning to learners in education (instruction)-oriented ways[1].
Since the 1980s, AIED research has adopted the
view of constructivism, which exists in cognitive sci-ence and learning scisci-ence. This view claims that knowl-edge related to learning should be constructed by learn-ers themselves, whereas it could be communicated in CAI and ITS. AIED research has adopted the learning theory of constructivism in developing systems classi-fied into ILE (Interactive Learning Environment), in which learners are provided with scaffolds to enable them to construct knowledge by themselves via interac-tion with the systems[17]. AIED researchers have explored such scaffolding methods that allow learners to use tools to manipulate information or objects using a trial and error approach[18].
Since the 1990s, social constructivism has emerged. This learning view claims that knowledge and skill should be acquired from collaborative interaction and communication between learners. AIED research has adopted learning theories in social constructivism to explore methods for scaffolding collaborative interac-tion and communicainterac-tion such as discussion and debate[19]. These scaffolding methods have been inte-grated into systems called CSCL (Computer-Supported Collaborative Learning). In contrast to CAI and ITS, ILE and CSCL are viewed as learner-centered systems.
As shown in Fig. 3, the system paradigm has shifted from being education-oriented to learning-ori-ented[20], and methods have been accordingly explored from adaption to scaffolding. Currently, learning-ori-ented methods and systems have become more relevant than education-oriented methods owing to the need for self-learning skills in the 21st century[21].
3.3 Influence of Learning Media
Research on learning and education using technology has also been influenced by the learning media used[3]. In particular, the technological progress of learning media has promoted the diversification of learning aspects and styles. Therefore, researchers in JSiSE are investigating new aspects and styles of learning and developing various methods and systems for learning.
Fig. 4 shows the diverse learning media used in this research field and their relationships with various learning aspects/styles[3]. At the initial stage of the research, computers and PC were mainly used as the learning media. These involve processing symbols and patterns as information, which is the key technology necessary for representing models, and for developing
methods and systems for learning support. Information processing is commonly embedded in all learning media. Computers and PCs represent the individuality of learning, which cannot be implemented in the class-room for mass education. This contributes to fostering research on stand-alone learning environments such as CAI, ITS, and ILE. PCs have been used as one of the major learning media up to now.
Next to computers and PCs, network technology such as the Internet and the Web have been used as learning media. These promote collaborativeness and connectedness in learning, which have driven research on CSCL, distance learning, and life-long learning. Recently, methods and systems for social learning with Social Networking Service (SNS) are being explored[22].
In addition, mobile and ubiquitous devices used for learning such as smart phone, wearable watch, and RFID tag enhance the distribution, mobility, and ubiqui-tousness of learning. These have promoted research on mobile learning, ubiquitous learning, and situated learn-ing. In particular, researchers have explored methods of providing learners with instructive support appropriate for the specific place and situation detected by the devices[3].
The learning media mentioned above can be viewed as an enhancer that extends the learning envi-ronments for learners. On the other hand, Virtual Reality (VR), Augmented Reality (AR), and tablet PCs can be used as learning media to support learning with bodily
sense and experience. These can improve the learners’ immersion in the cognitive process related to learning, contributing to augmenting their sense and experience and reinforcing their subjectivity in learning[23].
A humanoid robot can also be used to not only sup-port the cognitive aspect but also the emotional aspect of learning, such as engagement owing to its social pres-ence[24], [25]. A robot is viewed as embodied and anthro-pomorphized learning media[26]. It produces humanlike communication with learners. In particular, it has the potential to arouse emotions by means of its behav-ior[27].
For instance, we proposed a robot lecture, where a robot substitutes for human lecturers, and enhances their lecture[28]. In the robot lecture, the robot essentially reproduces the lecture, including the behavior necessary for the lecture. It also reconstructs the inappropriate behavior of human lecturers into appropriate behavior. The results of the case studies with the robot lecture sys-tem suggest that a robot can hold learners’ attention dur-ing its lecture usdur-ing non-verbal behaviors such as point-ing gesture and eye contact to keep them engaged in the lecture[28]. It is also suggested that the robot lecture pro-motes both learners’ engagement and their understand-ing of the lecture content[29].
These learning media can be considered as those that enhance emotion, bodily sense, and experience in addition to cognition. It would be challenging but inter-esting to explore methods and systems for these aspects, which has been recently initiated by JSiSE[26].
Besides the above-mentioned learning media, vari-ous sensors such as motion sensor and wearable sensor allow researchers to explore motor skill development. This is another research direction that has been consid-ered.
ICTs used for learning are increasingly diverse; however, the importance of understanding how learning occurs remains unchanged in terms of resolving the essential issue of how to support learning using ICTs. In particular, we believe that the design approach to mod-eling such diverse learning is promising a research theme in informatics for learning and education.
4. Future Directions
Finally, this paper describes the future directions on informatics for learning and education. Although most existing studies in this research field have focused on Figure 4. Diversity of learning media and learning aspects/
the cognitive aspects of learning to develop models, methods, and systems, it is essential to explore the emo-tional aspects of learning as these have significant influ-ence on the cognitive aspect[30]. Although research on affective computing has made attempts to model emo-tion[31], the JSiSE aims to further the use of emotion enhancers such as VR, AR, and robot to design models for the emotional aspects such as engagement and to develop methods for adapting to and scaffolding learn-ers’ emotion.
For instance, we have used a communication robot to design a model of engagement to be derived from interaction between learners and the robot, as shown in Fig. 5[26]. This model suggests the communication pat-terns for extracting the learners’ engagement while con-sidering factors that influence engagement. It also sug-gests the constituents of robot behavior necessary for forming communication patterns. We have referred to the model to develop some robot systems that can com-municate with learners so that they can be engaged in the contents communicated[28], [32].
In addition, we currently need to re-consider learn-ing by humans in comparison with machine learnlearn-ing, which has rapidly progressed as AI research. In particu-lar, we need to focus on the aspects of human learning that machine does not yet possess. The representative aspects are meta-cognition, creativity, self-directedness,
belief, etc., which are related to learning. Although these aspects have been explored in part by JSiSE, more attention should be paid to them in designing models, methods, and systems using new and suitable learning media.
5. Conclusion
Numerous ICTs and educational applications are cur-rently easily available, which can be used by researchers and practitioners to integrate learning environments and educational systems without careful consideration of how learners learn. In this approach, however, it becomes difficult to develop the methods and systems appropriate for the learning aspects the ICTs and educa-tional applications promote. The key to resolving the issue on how to support learning using technology is to understand learning.
This paper highlighted research on informatics for learning and education, which explores the key issue of how to model learning and develop methods and sys-tems for systematic learning support. In this paper, we emphasized the importance of the design approach to the theoretical development of models, methods, and systems, which allows researchers to handle diverse learning aspects owing to the technological progress of the learning media, and create suitable learning environ-Figure 5. Model of engagement to be derived from learner–robot interactions (modified from [26])
ments and educational systems.
This paper also described the JSiSE, which has been exploring informatics for learning and education. In particular, it not only encourages theoretical develop-ment, but also practical use of the models, methods, and systems developsed, which involves circular interaction between them. It is expected that JSiSE will continue to promote this interaction to foster this research field in the future.
Acknowledgements
This study was supported in part by the JSPS KAKENHI Grant Numbers 17H01992, 18K19836, and 20H04294.
References
[1] E. Wenger, Artificial Intelligence and Tutoring Systems:
Computational and Cognitive Approaches to the Communication of Knowledge. Morgan Kaufmann
Publishers Inc., 1987.
[2] J. Carbonell, “AI in CAI: An artificial-intelligence approach to computer-assisted instruction,” IEEE Trans.
Man-Machine Syst., vol. 11, no. 4, pp. 190–202, 1970.
[3] A. Kashihara, “Design of learning model,” (in Japanese)
J. Jpn. Soci. Artif. Intell., vol. 35, no. 2, pp. 201–207,
Mar. 2020.
[4] A. Kashihara, “Learning informatics as modeling of learning and learning environments,” (in Japanese), J.
Jpn. Soc. Arti. Intell., vol. 30, no. 4, pp. 473–476, Jul.
2015.
[5] JSiSE, Homepage. https://www.jsise.org (accessed Nov. 4, 2020).
[6] APSCE, Homepage. https://www.apsce.net (accessed Nov. 4, 2020).
[7] IAIED, Homepage. https://iaied.org (accessed Nov. 4, 2020).
[8] JSiSE, Transactions of Japanese Society for Information and Systems in Education. https://www.jstage.jst.go.jp/ browse/jsise/-char/ja/ (accessed Nov. 4, 2020).
[9] JSiSE, The Journal of Information and Systems in Education. https://www.jstage.jst.go.jp/browse/ejsise/ (accessed Nov. 4, 2020).
[10] LIGP, Homepage. http://ligp.gingerapp.co.jp (accessed Nov. 4, 2020).
[11] R. K. Sawyer, Ed., The Cambridge Handbook of the
Learning Sciences. Cambridge University Press, 2006.
[12] H. A. Simon, The Sciences of the Artificial. MIT Press,
1996.
[13] A. Kashihara and G. Shiota, “Knowledge construction with visual and Pseudo-Haptics,” in Proc. of the 12th Int.
Conf. Intell. Tutoring Syst. (ITS2014), LNCS 8474,
Springer, 2014, pp. 61–68.
[14] A. Kashihara and N. Akiyama, “Learning scenario cre-ation for promoting investigative learning on the Web,” J.
Inf. Sys. Edu., vol. 15, no. 1, pp. 62–72, 2017.
[15] H. Mandl and A. Lesgold Eds., Learning Issues in
Intelligent Tutoring Systems. Springer-Verlag, 1998.
[16] R. R. Burton, “Diagnosing bugs in a simple procedural skill,” in Intelligent Tutoring Systems, D. Sleeman and J. S. Brown, Eds., London, Academic Press, 1982, pp. 157– 183.
[17] B. P. Woolf, Building Intelligent Interactive Tutors:
Student-Centered Strategies for Revolutionizing E-learning. Morgan Kaufmann Publishers Inc., 2008.
[18] S. P. Lajoie, Computers as Cognitive Tools: No More
Walls. vol. 2, 2000, Lawrence Erlbaum Associates.
[19] M. Scardamalia and C. Bereiter, “Computer support for knowledge-building communities,” J. Learning Sci., vol. 3, no. 3, pp. 265–283, 1994.
[20] D. A. Norman, “Learner-centered education,” Commun.
ACM, vol. 39, no. 4, pp. 24–27, 1996.
[21] P. Griffin, B. MaGaw, and E. Care, Eds., Assessment and
Teaching of 21st Century Skills. Springer, 2011.
[22] S. Hasegawa and A. Kashihara, “Fundamental technologies for education/learning support in network community,” (in Japanese), Trans. Jpn. Soc. Inf. Syst. Educ., vol. 28, no. 1, pp. 9–20, Jan., 2011.
[23] J. Radianti, T. A. Majchrzak, J. Fromm, and I. Wohlgenannt, “A systematic review of immersive virtual reality applications for higher education: Design ele-ments, lessons learned, and research agenda,” Comput.
Educ., vol. 147, DOI: 10.1016/j.compedu.2019.103778,
2020.
[24] T. Belpaeme, J. Kennedy, A. Ramachandran, B. Scassellati, and F. Tanaka, “Social robots for education: A review,” J. Sci. Robotics, vol. 3, no. 21, DOI: 10.1126/sci-robotics.aat5954, 2018.
[25] I. Leite, G. Castellano, A. Pereira, C. Martinho, and A. Paiva, “Empathic robots for long-term interaction: Evaluating social presence, engagement and perceived support in children,” Int. J. Soc. Robotics, vol. 6, pp. 329– 341, 2014.
[26] A. Kashihara, “Research on learning with social robot,” (in Japanese), Trans. Jpn. Soc. Inf. Syst. Educ., vol. 37, no. 2, pp. 73–82, Apr., 2020.
[27] D. Szafir and B. Mutlu, “Pay attention!: Designing adap-tive agents that monitor and improve user engagement,” in Proc. 30th ACM Conf. Hum. Factors in Comput. Syst.
(CHI2012), pp. 11–20, 2012.
[28] T. Ishino, M. Goto, and A. Kashihara, “A robot for recon-structing presentation behavior in lecture,” in Proc. 6th Int.
Conf. Human-Agent Interaction (HAI2018), pp. 67–75,
2018.
[29] A. Kashihara, T. Ishino, and M. Goto, “Robot lecture for enhancing non-verbal behavior in lecture,” Proc. 20th Int.
Conf. Artif. Intell. (AIED2019), pp. 128–132, 2019.
[30] E. A. Skinner, T. A. Kindermann, J. P. Connell, and J. G. Wellborn, “Engagement and disaffection as organizational constructs in the dynamics of motivational development,” In Educational Psychology Handbook Series. Handbook
of Motivation at School, K. R. Wenzel, and A. Wigfield
Eds., Routledge/Taylor & Francis Group, 2009, pp. 223– 245.
[31] S. Carberry and F. de Rosis, “Introduction to special Issue on Affective modeling and adaptation,” J. User Modeling
and User-Adapted Interaction, vol. 18, no. 1, pp. 1–9,
2008.
[32] Y. Adachi and A. Kashihara, “A partner robot for promot-ing collaborative readpromot-ing,” Proc. Int. Conf. Smart Learnpromot-ing
Environ. (ICSLE2019), pp. 15–24, 2019.
Akihiro Kashihara received the B.Eng. and M.Eng. degrees in Computer Science and System Engineering from Tokushima University in 1987 and 1989, respectively, and the Ph.D. in Information and Computer Science from Osaka University in 1992. He was an Assistant/ Associate Professor in the Institute of Scientific and Industrial Research, Osaka University from 1992 to 2004. He is now a Professor in the Graduate School of Informatics and Engineering, The University of Electro-Communications. He is working on learning informatics. His main research interests lie in design-ing of learndesign-ing models, learndesign-ing with robot, scaffolddesign-ing with cognitive tools, and designing intelligent learning environ-ments.
![Figure 3. Transition of existing learning theories/views (modified from [3])](https://thumb-ap.123doks.com/thumbv2/123deta/6821651.1167538/5.773.128.646.153.521/figure-transition-existing-learning-theories-views-modified.webp)
