This interactive event targets at increasing the interaction among researchers who are interested in the applications of artificial intelligence in the education domain. Some of the potential contributions of artificial intelligence to educational activities had been realized in the past decades. Adaptive assessment of students' competence in natural languages, such as English, has been used in the real world. Two students who take the language tests do not have to solve the same set of test items. Enabling software to work with her users in a context-dependent manner has also started to show attractiveness.
In this interactive event, invited speakers who are adept at student modeling, virtual agents, and knowledge engineering will share their expertise with the audience. These topics contain a subset of core technologies of AIED, and cover important areas which bridge the research results and educational applications in the real world. The participating audience will have the chance to exchange opinions and share experiences during the formal panel discussion and the relaxing social activities. In the current plan, the panel discussion includes four long presentations, five short talks, and a recess. The social activities will allow the participants to get along with each other in a relaxed atmosphere, in the hope to nurture international cooperation in the future.
Schedule and Slides
1 December 2009 (panel discussion in morning and social activities in the afternoon, complete conference programme)
We thank the speakers for their sharing experiences with us. If you find that the slides are helpful for your research, please cite appropriate papers of the speakers.
Approximate Time | Speakers/Activities | Title/Abstract | Slide Files |
---|---|---|---|
0900-0925 | Riichiro Mizoguchi | link | mizoguchi.ppt |
0925-0935 | Shu-Chen Cheng | link | cheng.ppt |
0935-1000 | Antonija Mitrović(Tonja Mitrovic) | link | mitrovic.ppt |
1000-1010 | Chi-Jen Lin | link | cjlin.ppt |
1010-1015 | flexible... | ||
1015-1040 | break/interaction | ||
1040-1105 | Gautam Biswas | link | biswas.pptx |
1105-1115 | Zhi-Hong Chen | link | chen.pptx |
1115-1125 | Hao-Chiang Koong Lin | link | hclin.ppt |
1125-1150 | Tsukasa Hirashima | link | hirashima.ppt |
1150-1200 | Wing-Kwong Wong | link | wong.ppt |
1200-1230 | discussion | ||
1230----- | lunch and social activities |
This interactive event will include two major parts. This first part is a panel discussion, and the second part will include informal interactions, including social activities, after the panel discussion.
We invited four internationally renowned scholars to hold four about 25 minute talks. They are Gautam Biswas, Tsukasa Hirashima, Antonija Mitrović (Tonja Mitrovic), and Riichiro Mizoguchi. Five of the proposers of this interactive event will also participate in the panel discussion, and each will give a short talk of about 10 minutes. In addition to the presentations, we would like to allow participants of the interactive events to discuss with the panelists, and the discussion should take about between 20 to 30 minutes. Hence, the panel discussion will last about 180 minutes (four 25 minute talks, five 10 minute talks, and one 30 minute discussion). For a more reasonable and non-exhausting planning of time, we hope to include a recess of about 20 minutes, so this event will span about three hours and twenty minutes in total.
While this event consists of lectures as the core activities, the speakers will prepare extended abstracts for their talks. To help the participants receive more information about the lectures effectively, we would like to pass working notes to the participants on site.
After the formal panel discussion, we would like to have social activities, including a lunch with the invited speakers.
Speakers, chairperson, titles and abstracts
Over the past decade, new and exciting technologies have created opportunities for developing rich open-ended learning environments than combine different learning paradigms and resources. Students can complete quests in game environments, engage in inquiry, interact with virtual agents, run science simulations, take quizzes, access the web, and more generally make choices about different learning resources and activities. We believe that the choices students make are closely related to their learning and indicative of what they will be able to learn once they leave their scripted classroom learning environments. We will describe a new Choice-Adaptive Intelligent Learning Environment (CAILE) that combines multi-agent adaptive technologies and service architectures to provide a framework for designing extendible and reconfigurable learning environments. This approach extends our current research on Teachable Agents (TAs), where students learn by teaching an agent, and are provided with multiple opportunities (choices) to monitor their learning behaviors. We have applied data mining and sequence analysis schemes to student activities collected in log files to analyze their learning behaviors, from which we infer their use of strategies for learning. When students use suboptimal strategies, the intelligent agents can adapt and suggest alternative behavior choices that can improve learning. Ideally, by permitting student choice, while at the same time providing adaptive metacognitive support for making better choices, we can help students learn in the computer environment and transfer this learning to more real-world situations.
To make something is a promising way to understand it deeply. Several researches have suggested that to make problems is also effective approach to comprehend the problems and solution methods for them. We call this learning "learning by problem-posing". I have investigated interactive learning environment for learning by problem-posing. In this learning, learners make various types of problems including wrong problems. Therefore, to make the learning effective, assessment and feedback for the posed problems are indispensable. Self-assessment is usually difficult for learners. Teacher-assessment is the most reliable way, but it is impossible for a teacher to take care of individual learner and problem in usual teaching situation. Although peer-assessment is a practical way, it is very difficult to manage the assessment. I have investigated agent-assessment and realized several interactive environments for learning by problem-posing. I have also confirmed the learning effect in some of them at practical teaching situation. In this presentation, I will introduce the concept of "interactive environment for learning by problem-posing" and practice use of them.
One might have once dreamed if computer could understand learning/instructional theories. Recent advancement of ontological engineering has enabled the dream to come true. Since late 90's, we have tackled this problem with my colleagues and have come up with a theory-aware authoring system for each of individual learning and collaborative learning. The former is called SMARTIES and latter CHOCOLATO. While SMARTIES is based on a comprehensive ontology of learning and instructional theories, CHOCOLATO on ontology of collaborative learning. This talk discusses background philosophy behind the research as well as its technological details and usefullness of the systems.
Recently, educational agents have been attracting more and more research interests in the research filed of technology-enhanced learning (TEL). Two possible reasons explain why educational agents are worth investigating. The first one is that the advance of multimedia technology enables the believable behaviors of educational agents. The second reason is the recognition of potential benefits that educational agents bring. These potential benefits foster the investigation of educational agents in TEL.
However, for educational agents, we need to ask a foundational question: what is the main difference between educational agents and learning tools? We also tend to devise powerful tools for educational purposes. So far as this opinion is concerned, educational agents should not be merely powerful tools. That is, educational agents should have some core values different from educational tools. We belief one of the core values is—educational agents must also be human-like agents who can “care” students’ learning.
Therefore, in this study, we propose an approach, reciprocal caring approach (RCA), to integrating the two perspectives for designing educational agents. The RCA involves two kinds of “care-giver” roles, in which one is played by students and the other is played by educational agents. Students and educational agents can care each other. We attempt to explore deeper relationship established between students and educational agents. Based on the relationship, educational agents might have more opportunities to benefit students’ learning in both affective and cognitive aspects. My-Pet-My-Quest, a game-based educational agent system, is described in this work as an example to show how to design such educational agent by RCA.
This research has adopted artificial intelligence technology to build up a Plant-Enquiry system to enhance the user’s cognition about the plants. The concept of Web 2.0 has been used to construct a share-learning network. Doing on-line and real time editing of the digital multimedia could deliver the message in a more effective and more convenient way, and, by implementing the technology of video-film indexing and marking, it allows the users to more effectively access huge amount of useful information; the technology implementation of GPS and QR-Code has been added to provide the course-editing function of M-Learning. Combining the conveniences mentioned and the fast operations, it achieves plant-retrieving, complete plant-introduction, the building up and collection of plant database, the integrated application of GPS, QR-Code and map information; it has also reduce the development difficulty of digital course design and the cost.
In order to provide tutoring service by computers, most system developers must implement knowledge models of related subject domains. The development of knowledge models is known to be difficult and expensive, and usually requires knowledge engineering expertise, thus most of human tutors are unable to perform it. Contrarily, human tutors are excellent at interacting with tutees. Thus, accumulating interaction data is an excellent approach for human tutors to create computer tutors. In this paper, we describe a data model and an approach to providing tutoring service by accumulating human tutoring interaction data.
This paper is aimed at implementing a system to help students concentrate in class. The system consists of three main stages: speech analysis, emotion analysis, and avatar output. First, the speech analysis transcribes vocal captures to text, which in turn is consumed by the emotion analysis. A number corresponding to the analyzed emotion in outputted to the avatar. Based on the emotion, the avatar makes a dialogue choice for interacting with the student. Experiment results shows that the system is promising and has a lot of potential.
This paper presents a tool for drawing dynamic geometric figures by understanding the texts of geometry problems. With the tool, teachers and students can construct the dynamic geometric figures on a webpage by inputting the problem in natural language. First we need to build the knowledge base for understanding geometry problems. With the help of the knowledge base engine InfoMap, geometric concepts are extracted from an input text. The concepts are then used to output a multi-step JavaSketchpad script, which constructs the dynamic geometry figure on a webpage. Finally, the system outputs the script as an HTML document that can be read and visualized with an internet browser. Furthermore, a preliminary evaluation of the tool shows that it produced correct dynamic geometric figures for over 88% of the problems from textbooks.