1. Intelligent Tutoring Systems
Shekh Zaid Mohammad
University Of Haifa, Information Systems Department
Abstract: This paper introduces Intelligent Tutoring Systems (ITS), ITS are computerized learning environments that incorporate computational models from the cognitive sciences, learning sciences, computational linguistics, artificial intelligence, mathematics, and other fields. An ITS tracks the psychological states of learners in fine detail, a process called student modelling. The psychological states may include subject matter knowledge, skills, strategies, motivation, emotions, and other student attributes
Keywords: Intelligent Tutoring Systems (ITS), student model, domain model, cognitive science.
1. INTRODUCTION
Intelligent Tutoring Systems (ITSs) are computer-based instructional systems with models of instructional content that specify what to teach, and teaching strategies that specify how to teach. They make inferences about a student’s mastery of topics or tasks in order to dynamically adapt the content or style of instruction. Content models (or knowledge bases, or expert systems, or simulations) give ITSs depth so that students can "learn by doing" in realistic and meaningful contexts. Instructional models allow the computer tutor to
more closely approach the benefits of individualized instruction by a competent pedagogue.
The systems have transformed to become a true intersection of computer science, cognitive psychology and educational research. They have offered various focus points in system development across various time periods highlighting research areas, on and off being addressed by researchers from time to time.
This includes a description of the computational mechanisms of each type of ITS and available empirical assessments of the impact of these systems on learning gains.
It is important to treat the effect of human tutoring, including the practice of human tutoring, relevant pedagogical theories and this because some ITS are built with the
2. guidance of what human tutors know about effective pedagogy , this paper identifies some future for the ITS field .
2. HISTORY
The history of ITS has always had a cross fertilization with the psychological sciences in addition to implementing intelligent computational mechanisms .There was the vision that learning could substantially improve by developing adaptive learning environments that took advantage of the latest advances in intelligent systems. Some of these ITS attempted to model characteristics of human tutoring whereas others adopted ideal, rational models of intelligent systems.
There has been evolution from a very primitive form of computer-assisted instruction, ranging through various forms of e-learning systems, progressing to form learner adaptive systems, to modern day ITS.
In developing an Intelligent Tutor, or ITS , the goal has been to capture at least some of the impressive increase in learning -- an average of 2 sigma in the studies by Bloom (1984) -- that tends to result from one-on-one tutoring as compared to group instruction. The much quoted "2-sigma effect" of having an individual tutor meant that the average student who had been tutored performed better than 98% of the students who had been taught in the classroom. In comparing the effect of one-on-one tutoring with the effect of a number of other variables and treatments, Bloom found that none of the others produced an effect as large as 2 sigma.
3. I can summarize the main goals of ITS and shortened at several points, ITS:
To modularize the curriculum.
To engage the students in sustained reasoning activity .
To interact with the student based on a deep understanding of the students behavior.
Collect data which instructors could use to tutor and remediate students.
Fig. 1. Typical Architecture of an ITS
3. STRUCTURE OF ITS
A typical ITS, has the following four basic components:
The Domain Model
The Student Model
The Tutoring model
The User interface Model
3.1 The Domain Model : The domain model , also known as the cognitive model , consists the facts, rules , and problem solving strategies of the domain context like algebra, geometry, and programming languages . It serves as a source of expert knowledge, a standard for evaluation of the student’s performance and diagnosis of errors.
This module is at the heart of an intelligent tutoring system and provides the basis for interpreting student actions.
4. 3.2 The Student Model : The student model is an overlay on the domain model. It emphasizes cognitive and affective states of the student in relation to their evolution as the learning process advances. As the student works step-by- step through their problem solving process, the system engages itself in model tracing process. Anytime there is any deviation from the predefined model, the system flags it as an error. Students can increase the accuracy of ITS more by indicates their feedback about every step of the problem solving process. connections between complex learning and emotions have received increasing attention in the fields of psychology and education , Studies that have tracked the emotions during tutoring have identified the predominate emotions, namely confusion, frustration, boredom, anxiety, and flow/engagement, with delight and surprise occurring less frequently. This model encourages the tutor to nurture the student by being empathetic and attentive to the student’s needs, to assign tasks that are not too easy or too difficult, to give indirect feedback on erroneous student contributions rather than harsh feedback.
3.3 The Tutoring model : The tutor model accepts information from the domain and student models and devices tutoring strategies with actions.
This model regulates instructional interactions with student. makes use of knowledge about the student and its own tutorial goal structure. It tracks the learner's progress, builds a profile of strengths and weaknesses relative to the production rules (termed as ‘knowledge-tracing’).
3.4 The User interface Model : This is the interacting front-end of the ITS. It integrates all types of information needed to interact with learner, through graphics, text, multi-media, key-board, mouse-driven menus .Prime factors for user-acceptance are user-friendliness and presentation.
The ITS problem solving environment should approximate the real world problem solving environment.
The ITS problem solving environment should facilitate the learning process.
The tutor – system – make the decisions based on, first of all, the domain model that represent the relevant knowledge that been thought, consists concepts and principles.
5. We need the pedagogical model knowledge for Teaching strategies, remediation , curriculum . And also needing any given points information about student interaction (what is the student knowledge? ,what the students believes) .
based on this information the tutor can generate a tutorial actions witch being the beginning interaction activity. But the system should understanding what is the good way or how to present the same activity to different students that they may have different knowledge, cognitive thinking, how to see information activity, by text or images or anything else.
Furthermore, when activity selected the system does need have the answer, there is a component that able to used the representations in the target domain knowledge to generate on the fly solutions , then the computer solution and the student solution are matched using the student modeler to update systems believes (what the student knows? What the student do? Fail?) or any relevant information , and this information go back in the cycle to represent hints , feedback, corrections to help students continues their problem solution.
Fig. 2. Ideal ITS
6. 4. ITS TYPES AND TECHNIQUES
4.1 Cognitive tutor: One of the salient success stories of transferring science to useful technology is captured in the Cognitive Tutor. In the Cognitive Tutor paradigm, the tutor has a psychological model capable of solving the problems presented to the student, A production rule is an “IF [state S], THEN [action A]” expression that designates that action A is likely to be performed when a state S is perceived and/or in working memory, An action may be a cognitive act or a behavior. Over the course of problem solving, production rules - production rules accumulate in long-term memory, that get activated by the contents of working memory, and that are dynamically composed in sequence when a particular problem is solved - in long-term memory are activated and capture the student’s knowledge and mastery of the skill. The correspondence between the student's behavior and production rule firing is known as model-tracing.
Knowledge-tracing assigns a probability of student mastery to each correct production rule based on student behavior.
When the system has decided that enough of the skill requirements have been met for mastery, the tutor and student move on to a new section.
One such system is Andes (Fig. 3), a coached learning environment for classical physics that has been in development since 1996 by researchers at the Learning Research and Development Center of the University of Pittsburgh and at the United States Naval Academy. This intelligent tutor allows students to solve physics problems in an environment that provides visualization, immediate feedback, and procedural and conceptual help. It runs under both the Windows 98 and NT operating systems.
Andes consists of an authoring module, developed by the Naval Academy, and a student module, developed by the Learning Research and Development Center. This software is still under development. It currently tutors students in the areas of static forces, translational and rotational kinematics, translational and rotational dynamics, energy, and linear and angular momentum.
7. Fig. 3. A typical one-dimensional problem solved via components.
4.2 Constrain –based Tutor: The core idea of CBM is to model the declarative structure of a good solution rather than the procedural steps leading to a good solution, In CBM, the declarative structure of a good solution is composed of a set of state constraints (Ohlsson, 1994). Each state constraint is composed of a relevance condition (R) and a satisfaction condition (S).
CBM is allegedly able to account for a wider array of student behavior because it can accommodate greater deviations from the correct solution path. Because solution paths are not explicit in CBM, feedback is triggered by violation of satisfaction constraints. Moreover, the content of the feedback is not tied directly to a solution step but rather to a global constraint in the solution space. The implications of constraint-based student models have been explored in the literature, leading to a more elaborate and nuanced understanding of the tradeoffs between CBM and other modeling approaches.
KERMIT, is an entity-relationship tutor that focuses on database design. Constraints in KERMIT represent database design principles. students using KERMIT had significantly higher learning gains than students from the same class who used KERMIT without constraint-based feedback, with an effect size of 0.63.
8. 4.3 Case- based Reasoning: ITS with Case-based reasoning (CBR) are inspired by cognitive theories in psychology, education, and computer science that emphasize the importance of specific cases, exemplars, or scenarios in constraining and guiding our reasoning.
problem solving makes use of previously encountered cases rather than by proceeding from first principles.
systems generally follow four steps:
RETRIEVE the most similar case(s), as indexed in the memory system
REUSE the case(s) to attempt to solve the current problem
REVISE the proposed solution to the current problem if necessary, and RETAIN the new solution as a part of a new case that is stored in the memory system.
learning in the CBR paradigm goes beyond adding new cases after successful REUSE and beyond adding new cases after failed REUSE and successful REVISION. CATO, which was designed to help beginning law students acquire basic argumentation skills. CATO uses the model for a number of purposes, including the dynamic generation of argumentation examples. In a second evaluation study, carried out in the context of an actual legal writing course, we compared instruction with CATO against the best traditional legal writing instruction. The results indicate that CATO's example-based instructional approach is effective in teaching basic argumentation skills. However, a more “integrated” approach appears to be needed if students are to achieve better transfer of these skills to more complex contexts.
4.4 Conversational Agents: Human Society has used conversation as an efficient, reliable and adaptive means of exchanging knowledge. One of the earliest grand challenges of artificial intelligence has been to extend this conversational medium to interactions with automated participants like computers or robots. We refer to these automated participants as Conversational Agents.
9. 5. FUTURE ASPECTS AND CONCLUSION
The technology for intelligent tutoring systems appears to be poised for wider use. One possibility is for tutors to be programmed, where the conversational assistance the tutors offer can provide students a sense of instructor interaction that might otherwise be missing. As more students use intelligent tutoring systems, greater amounts of data retrieved from student input can be used to design improvements in the systems themselves. Also, some researchers are working with voice recognition and simulated natural human dialogue to improve the feedback cycle between student and tutor. In this way, intelligent systems might one day be able to respond not just to a student’s words but also to tone, facial expression, or body language.
Personalization of learning environments has significant impact in learning process improvement. This paper has presented a models that combine learning styles and emotion as two factors to improve this process. We can improve our models with considering emotions like shame, anger and stress and proposing solutions to omit these emotions during learning process. chooses the learning style which has the highest point in the questionnaire, but in the future we can consider multimodal learning styles that are a combination of two or more styles.
One interesting subarea of ITS research is dialogue-based tutoring systems, which can be considered an elaboration of the architecture described above. The underlying metaphor of these systems is conversation, often in typed form to avoid the specialized algorithms needed for speech recognition, and often including non-verbal acts such as pointing or graphics display in addition to speech acts. These systems extend research on domain independence to include issues of language independence, i.e. the extent to which resources developed for natural language understanding and generation can be reused in different tasks and domains.
There is a need for ITS
researchers to explore the possibility and develop a framework for integrating social networking agents. There is a need to drill down and analyze emotional states of the learner and accordingly align the focus as well as learning/teaching strategies of ITS. There is immense future research direction embedded in it. With an increased demand for portable devices, various hand-held intelligent tutoring systems promise rich dividend.