Intelligent tutoring system

About: Intelligent tutoring system is a research topic. Over the lifetime, 3472 publications have been published within this topic receiving 58217 citations. The topic is also known as: ITS.

A bibliometric analysis of worldwide educational artificial intelligence research development in recent twenty years

Abstract: Educational artificial intelligence (EAI) refers to the use of artificial intelligence (AI) to support personalized and automated feedback and guidance in the educational field. Inevitably, it serves as a more important part of the educational system in the coming years. However, novel development in this field has been inadequately reviewed and conceptualized in a visualized, objective and comprehensive way. In this view, a bibliometric analysis was conducted to obtain an overview of its trends from publication outputs, countries’ cooperation, cluster analysis, and research evolution. Around 8660 Scopus-published articles from 2000 to 2019 were gathered for analysis using CiteSpace and Alluvial generator. In the study, a growing interest in EAI research and deepening cooperation among countries was first identified, entailing favorable conditions for promoting globalization in this aspect. Afterward, five core clusters were established for the intellectual structure of EAI, including intelligent tutoring system, learning system, student, labeled training data, and pedagogy. The development of EAI research was further conceptualized as follows: (a) technological foundation; (b) technological breakthrough; (c) intelligent application; and (d) symbiotic integration. Finally, three prospective directions for future EAI research were suggested.

Design framework of adaptive intelligent tutoring systems

Abstract: The aim of this study is to examine adaptation elements and Intelligent Tutoring System (ITS) elements used in Adaptive Intelligent Tutoring Systems (AITSs), using meta-synthesis methods to analyze the results of previous research. Toward this end, articles appearing in the Web of Science, Google Scholar, Eric and Science Direct databases in 2000 and later were identified with the keyphrase “adaptive intelligent tutoring system.” Application of exclusion and inclusion procedures to the articles accessed in the search resulted in the selection of 32 articles, which were analyzed using meta-synthesis methods and then evaluated in the light of prespecified themes and elements used in AITSs were determined. According to results, the systems were designed for a wide range of fields such as Information Technologies, Mathematics, Science, Medicine, and Foreign Language Education. In these systems, content adaptation was generally used, based mostly on such criteria as feedback, student level, student learning and cognitive styles, and student performance. And besides 4 basic ITS modules (knowledge, student, teaching and user interface), some different modules such as guide module, strategy module, personal learning module, knowledge base module, communication module, system administrator module and messaging module were used. Finally, some suggestions were given for such studies in the future.

Does It Really Matter whether Students' Contributions Are Spoken versus Typed in an Intelligent Tutoring System with Natural Language?

Abstract: There is the question of whether learning differs when students speak versus type their responses when interacting with intelligent tutoring systems with natural language dialogues. Theoretical bases exist for three contrasting hypotheses. The speech facilitation hypothesis predicts that spoken input will increase learning, whereas the text facilitation hypothesis predicts typed input will be superior. The modality equivalence hypothesis claims that learning gains will be equivalent. Previous experiments that tested these hypotheses were confounded by automated speech recognition systems with substantial error rates that were detected by learners. We addressed this concern in two experiments via a Wizard of Oz procedure, where a human intercepted the learner's speech and transcribed the utterances before submitting them to the tutor. The overall pattern of the results supported the following conclusions: (1) learning gains associated with spoken and typed input were on par and quantitatively higher than a no-intervention control, (2) participants' evaluations of the session were not influenced by modality, and (3) there were no modality effects associated with differences in prior knowledge and typing proficiency. Although the results generally support the modality equivalence hypothesis, highly motivated learners reported lower cognitive load and demonstrated increased learning when typing compared with speaking. We discuss the implications of our findings for intelligent tutoring systems that can support typed and spoken input.

Evaluating the Effectiveness of a Cognitive Tutor for Fundamental Physics Concepts

Abstract: Evaluating the Effectiveness of a Cognitive Tutor for Fundamental Physics Concepts Patricia L. Albacete (albacete@isp.pitt.edu) Intelligent Systems Program; 607 Dixie Drive Pittsburgh, PA 15235 USA Kurt A. VanLehn (Vanlehn@cs.pitt.edu) Learning, Research and Development Center; University of Pittsburgh Pittsburgh, PA 15260 USA Abstract In this article we describe and analyze the evaluation of the Conceptual Helper, an intelligent tutoring system that uses a unique cognitive approach to teaching qualitative physics. The results of the evaluation are encouraging and suggest that the proposed methodology can be effective in performing its task. Introduction Several studies (e.g. Hake, 1998; Halloun & Hestenes, 1985a, 1985b) have revealed that solving physics problems of a qualitative nature, such as the one presented in figure 1, pose a great cognitive challenge for most students taking elementary mechanics classes. They uncover naive conceptions that are seldom removed or modified while completing their courses. Several attempts have been made to improve this situation though none has met with great success (Hake, 1998). Given that mechanics is a required course for most science majors, there is a clear need to improve its instruction. Toward this end we developed an intelligent tutoring system called the Conceptual Helper that follows a cognitive teaching strategy which is deployed emulating effective human tutoring techniques as well as successful pedagogical techniques and less cognitive demanding methods (Albacete, 1999; Albacete & VanLehn, 2000). In this article we describe the evaluation of the system and discuss its implications. Two steel balls, one of which weights twice as much as the other, roll off of a horizontal table with the same speeds. In this situation: a) both balls impact the floor at approximately the same horizontal distance from the base of the table. b)the heavier ball impacts the floor closer to the base of the table than does the lighter. c) the lighter ball impacts the floor closer to the base of the table than does the heavier. Figure 1. Example of a qualitative problem Brief description of the Conceptual Helper The Conceptual Helper is an intelligent tutoring system (ITS) designed to coach students through physics homework problem solving of a qualitative nature, i.e., those problems that do not require the use algebraic manipulation to be solved but so require the application of conceptual knowledge. The tutor is basically a model-tracing ITS enhanced by the use of probabilistic assessment to guide the remediation. As a model-tracing ITS it contains a cognitive model that is capable of correctly solving any problem assigned to the student. Model tracing consists of matching every problem-solving action taken by the student with the steps of the expert’s solution model of the problem being solved. This matching is used as the basis for providing immediate feedback to students as they progress through the problem. The system also has a student model which is represented by a Bayesian network. Each node in the network represents a piece of conceptual knowledge that the student is expected to learn or a misconception that the tutor can help remedy. Each node has a number attached to it that indicates the probability that the student will apply the piece of knowledge when it is applicable. As the student solves a problem, the probabilities are updated according to the actions taken by the student. The challenge for the tutor is to decide when to intervene and what to say when it does so. This task is particularly challenging in this domain because tutoring of qualitative knowledge usually takes the form of verbal discussions, which given the state of the art of natural language processing is not an option for the computer tutor. To take care of the issue of when to intervene, we emulated human tutors in two ways: first, by giving immediate feedback (red for incorrect; green for correct) on each student entry (Merrill et al., 1992) and second, by helping the student with post-problem reflection (Katz & Lesgold, 1994; Katz et al., 1996). However, most of our work went into the second issue—deciding what to say when intervening. Novel approaches were developed in three areas: 1) the teaching strategy, 2) the manner in which the knowledge is deployed, and 3) the way in which misconceptions are handled.

Facilitating deep learning through self-explanation in an open-ended domain

Abstract: Self-explanation has been used successfully in teaching Mathematics and Physics to facilitate deep learning. We are interested in investigating whether self-explanation can be used in an open-ended, ill-structured domain. For this purpose, we enhanced KERMIT, an intelligent tutoring system that teaches conceptual database design. The resulting system, KERMIT-SE, supports self-explanation by engaging students in tutorial dialogues when their solutions are erroneous. The results of an evaluation study indicate that self-explanation leads to improved performance in both conceptual and procedural knowledge.