Building a new student model to support adaptive tutoring in a natural language dialogue system
5
Citation
53
Reference
20
Related Paper
Abstract:
Intelligent computers are perfect for educational use since computers never get tired, they are available all the time, and they are accessible by users all over the world. One key component that makes an Intelligent Tutoring System more intelligent and more adaptive to individual student's needs is a student model.
This dissertation attacks problems of student modeling in a natural language based Intelligent Tutoring System—Circsim-Tutor. A student model that is composed of different sub-models has been built by considering the actual demands of different decision-making components of the system. We also discuss how to determine the knowledge required in each model by considering constraints imposed by the system. By comparing the quality of tutorial dialogues based on two versions of the student model in CIRCSIM-Tutor, we show the advantages of using the new student model.
We have also developed a new model of hinting for Circsim-Tutor. It can systematically deliver pedagogically sound and conversationally coherent hints in the tutoring dialogue by considering the domain knowledge, the type of error made by the student, the focus of the tutor's question, and the conversational history. In addition we have studied how to use machine learning techniques to discover tutoring rules from human tutoring transcripts.Keywords:
TUTOR
Intelligent tutoring system
Component (thermodynamics)
Cite
Tutormap is a tool that extends the mathematical computer system Maple with educational facilities for tutoring and assessment. It allows the student to test his/her understanding and mastering of both the Mathematics presented and the Maple language. It also provides the teacher with means to prepare tests and automatically grade them. This mathematics tutor is designed to operate in a networked environment, so that students can work independently on their assignments. We present and discuss the working version of Tutormap, its future extensions and our experience with it.
TUTOR
Cite
Citations (0)
TUTOR
Intelligent tutoring system
Course (navigation)
Cite
Citations (7)
Earlier work on Intelligent Tutoring Systems (ITSs) for programming focused more on teaching programming syntax than its application. The main tutoring approach is to present a problem specification for the student to solve, followed by intelligent analysis of the solution with various feedback. It is also observed that existing ITSs suffer from static domain knowledge and are restricted to the tutoring session. Therefore, this research proposes the development of a web-based ITS for both curriculum planners and implementer-tutors to teach students the application of the C++ Standard Template Library (STL) to problem solving.
From experience, it is discovered that students find the C++ STL difficult due to their weaknesses in understanding various object-oriented concepts. This ITS overcomes the learning and teaching challenges by modelling the program specification based on prerequisite concepts. Bayesian Theorem is applied to model the student’s knowledge and direct the tutoring intelligently. Bayesian probability reasoning is a well-known Artificial Intelligence technique for uncertainties management. The development of the C++ STL ITS applies practices from the eXtreme Programming methodology and J2EE technologies. The 3-tier architecture ITS constitutes three main components – Student Modelling Module, Tutoring Module and Users Administration Module providing the authoring of the domain knowledge dynamically. Hence, tutors can then fully participate in the design of the curriculum and tutoring sessions as well as in the implementation of the tutorials for their students for effective teaching and learning.
Both summative and formative evaluations were conducted on the C++ STL ITS. The evaluation results revealed that the Bayesian Theorem has the capability of modelling the student’s prerequisite and directing the student during the tutorial session. The Fuzzy Stereotyping of Students Expert System works well in categorizing the students according to four stereotypes – novice, beginner, intermediate and advanced.
Short term future enhancements include extending the tutorial questions, domain knowledge, accommodating more feedback on the programming syntax, and incorporating the fuzzy expert system into the C++ STL ITS. Three areas of research proposed for long term are application of alternative knowledge acquisition techniques, integration of learning styles into the student model, and representation of domain knowledge using ontologies.
Intelligent tutoring system
Cite
Citations (0)
Cite
Citations (287)
Abstract : The North Carolina A&T State University algebra tutoring dialogue project collects and analyzes algebra tutoring dialogues with the aim of describing tutoring strategies and language with enough rigor that they may be evaluated and incorporated in machine tutoring. The tutoring is computer-mediated chat-style, with the tutor and student collaboratively solving a problem and editing equations. Transcripts of these dialogues are annotated for tutorial intentions. language, and behaviors. Comparison of tutoring by more and less expert tutors allows us to focus on the strategies that are most effective. The most prominent we observe is a metacognitive strategy where the tutor tries to elicit from the student a categorization of the problem under discussion and to name a solution method. We combine these strategies into a structured set of tutorial goals, with example sentences for each. In subsequent tutoring the computer then suggests these goals and sentences to the - human tutor, who can pick use them in the dialogue. Measuring how much the tutors follow the computer-assistance, as well as student learning gains, allows to evaluate the validity of our structural analysis of tutoring.
Intelligent tutoring system
Cite
Citations (0)
Undergraduate students often struggle to learn optimal logic proof solving strategies in Discrete Math courses, primarily because of the open-ended nature of the domain. Students can, therefore, benefit from personalized tutoring, where they can receive user-adaptive support. Over the past decade, the advancements in the field of intelligent tutoring systems (ITSs) have made it possible to provide personalized tutoring with minimal involvement of a teacher or a human expert. While such tutoring systems have the potential to augment student learning on a large scale, few intelligent tutors are made open source. Deep Thought is a logic tutor where students practice constructing deductive logic proofs. Extensive research has been conducted for 11 years to provide data-driven intelligent tutoring support in Deep Thought. The logic tutor provides adaptive support using data-driven approaches on two levels: problem level, where the tutor decides whether the student should view the next problem as a worked example or they should solve it, and step level, where the tutor decides when an unsolicited partially-worked step should be provided to the student to direct them towards optimal problem-solving strategies. We have found encouraging evidence to support that the intelligent policies in Deep Thought help undergraduate students learn logic. Deep Thought is currently being used in discrete math classes at two universities: North Carolina State University, and the University of North Carolina at Charlotte. Our aim is to make this tutor available to a larger audience so as to contribute to the Computer Science Education community.
TUTOR
Intelligent tutoring system
Cite
Citations (2)
Abstract : With the support of the Cognitive Science Program of ONR, we are developing the capability to generate complex natural language tutorial dialogues for an intelligent tutoring system designed to help medical students understand the functioning of the negative feedback system that regulates blood pressure in the human body. We are convinced that a real natural language interface is vital to a tutoring system trying to help students learn complex concepts like negative feedback. The text generator for a tutoring system must be able to ask questions and provide hints in addition to generating definitions, descriptions, and explanations of the functioning of the physiological system and the underlying anatomy. The language understanding component must be ready to accept student responses and student initiatives, which are full of ellipses, novel abbreviations, wild spellings, and wilder grammar. Our work is embodied in a system called Circsim Tutor. Briefly described, Circsim Tutor presents the student with a set of clinical problems each of which results in a perturbation of blood pressure, and asks the student to explain step by step how the blood pressure is perturbed and how the perturbation is physiologically compensated for. The system conducts a tutorial dialogue in English, as the student does this. with the session organized around the students's errors in making these predictions.
Cite
Citations (0)
Computer based instructional systems either direct students so modelling their actions is tractable, or provide them with total autonomy, but give little support to learning and problem solving processes. Instructional principles for empowering the student are emerging whereby more of the responsibility of diagnosis and goal-setting is placed on the student. Critical to this view is providing an environment which makes the ramifications of students' actions clear so students can meaningfully assess their own performance. In the domain of word algebra, the meaning of formal expressions can be reflected in computer animation which depicts the corresponding situation. An unintelligent tutor — knowing nothing of the problem being solved and possessing no student model — helps students to understand problems and debug formal expressions.
TUTOR
Nothing
Cite
Citations (8)
TUTOR
Intelligent tutoring system
Presentation (obstetrics)
Cite
Citations (32)