Paradigm of instructional agent system Hsuanchao Huang & Reuben Edwards Department of Communication Systems Lancaster University E-mail:
[email protected],
[email protected]
Abstract- The UNIVERSAL Exchange for Higher Education will demonstrate the feasibility of an open exchange system for course units between institutions of higher education across the world. .The aim of this project is to develop and validate a model and standards that could later be widened to embrace other groups of higher education institutions and that could be transplanted into the market for training in industry, commerce and government. This paper aims to define the classifications for intelligent agents to investigate and classified a result of user’s learning style. The system will reference that result provide essential facilities available for user. In addition, intelligent agent could be helped user to probe them further then motivate self- learning. Users will thus benefits from a wider choice of facility units and essential helps within a system and from intelligent agents’. INTRODUCTION Thousands [1] of web-based courses have been created on the web over the last few years and intelligence agent technology provides exciting possibilities for the computer-based learning environment. Using intelligent agents to simulate instruction can serve as a powerful research tool to investigate teaching and learning. However, recent research in intelligent agent-based learning environments tends to focus on system development and implementation [2] rather than controlled user’s assessment of learning effectiveness. Designing an on-line course system should consider both factors and attributes between users and system. It would be reduced failure factors [3]. The main purpose of the system is to help learners to make better learning choices. We should begin by evaluated user learning effectiveness rather than focusing system development. In this paper we should define learners within an aware psychological view. First we should provide an overview of the intelligences and measured learners by uses of an intelligence testing, for defining different learning styles. Secondly, we shall describe the concept of quantum learning before dividing learners within restricted classifications. Within classified groups, users
ISBN: 1-9025-6009-4 © 2003 PGNet
could be provided with units guard forwards their learning type. Finally, we shall show some assessments on learning classifications for a wide range of users, studying at Lancaster University. 2. DEFINING LEARNERS There are many different way to define learner [4]. Some researchers defined user in degree levels [5] that are novice, advanced beginner, competence, proficiency and expertise. However thus is not effective to define the learner’s natural abilities and personality. We consider natural intelligences further than measure and classified them. Thus will more approached to our goal. 2.1 Overview intelligences Within a hundred year, some people have tried to probe the intelligences [6] until recent years just Defined the term intelligence is that a person’s ability to learn and memorize information, to recognize concepts and their relations, and to apply the information to their own behavior in an adaptive way [7]. But thus definition could be illustrated in general intelligence only. Is that the number of abilities can be examine? Gardner’s multiple intelligences theory denoted a personal contain eight intelligences [8] that argue intelligence falls into seven categories: musical cover the potential to distinguish and compose musical pitches, tones, and rhythms; mathematical/logical is ability to detect patterns, reason deductively and think logically; linguistic includes the capacity to effectively manipulate language to communicate oneself rhetorically or poetically; visual/spatial possesses ability to manipulate and create mental images in order to solve problems; Kinesthetic includes the type of skill; naturalist, personal intelligence includes awareness of one’s own feelings(intrapersonal) and the ability to notice individual differences in other people and to respond appropriately to them(interpersonal). 2.2 intelligences Test With Gardner’s multiple intelligences theory, some researchers have designed a test [9] for learners to probe
their intelligences. Principle of the test is that designed balance questions for within intelligence and by setting random orders, each answer for a question set five degrees, calculate a result of scores of each intelligence when a learner answer questions completely. Thus the result shows degree of containing intelligences that learner within test. 2.2 Quantum Learning Main purpose of Quantum Learning includes important aspects of neurolinguistic programming (NLP) [10] and is to study of how the brain organizes information, probes the relationship between language and behavior, further then classified in four different thinking that defined concrete sequential(CS) , abstract sequential(AS), abstract random(AR) and concrete random(CR), and can be used to create rap pore between learners and instructor , premise of it is that suggestion an and does affect the outcome of the learning situation and could be used almost interchangeably with suggest logy is accelerated learning to produce an effective learning experience [11]. 2.3 Classified learners As the test defined learners contained intelligences, we further then divided them within groups of thinkers. Such as describing of concrete sequential groups’ thinkers, we can establish that included mathematical/logical, musical, kinesthetic, visual/spatial intelligence. It is described fellows; this kind of thinker is based in reality and process information in an ordered, sequential, linear fashion. They need to organize tasks into step-by-step processes and strive for perfection at each step of the way (organized, precise, cautious, hardworking, perfectionist, orderly, completing). They like specific direction and procedures (planner wants directions, getting to the point). CS thinker reality consists of what they can detect through their physical senses of sight, touch, sound, taste, and smell (realistic). They notice and recall details easily and remember facts, specific information, formulas, and rules with ease (memorize). “Hand on is a good way for these thinkers to learn (practical, doing). 3. INTELLIGENT AGENTS IN INSTRUCTIONAL ENVIRONMENT Within instructional environment agent could be cognitive tools and intelligent tutors [12]. According to achieve functions of these two roles, we design agents to be a cognitive user tools belong to system and providing essential help when the user requested them.
3.1 instructional agent’s resebonsbilities In our system, agents responsibilities following. § § § §
should
achieve
four
Helping learner effective to find his/her personal learning styles. Navigating learner to motivate him/her. Watching and evaluating learner’s attitude anytime. Organizing information (Topics) and course for learner.
As we know learner’s intelligence can be probed by intelligence test. Generally, that test can be provided by a text type paper. However some learner found the more complicated to understand meaning and answer the questions. Thus the problems maybe affect the result of each item further than divided them to an unsuitable classification group. Doing the test with helping by mentor when they were misunderstanding, for the answers will be accuracy enhanced. As it can be known the agent mentor is more important for helping learners in this situation. The role of agent mentor is to explain and guide the user thus how to answer the questions. Another goal of agent is to transfer user’s attributes, referenced by system. In other words, agents could be sensors belong to system. Thus system accords with data provide facility units and will be able to record learners’ profiles further than provide parameters for course designers. 3.2 learning classification of users studying at Lancaster University At the beginning of term, we surveyed 60 students in three different departments by random selecting. We found that averages of student’s intelligences between departments are balanced (17.16826923, 16.8375, and 16.78571429). Students who are doing communication systems own higher average value of visual and spatial intelligence (19.57) than the other. Comparing with the other departments, students who are studying at Art department own lower average value of naturalist intelligence (12.5). CONCLUSION AND FURTHER WORK In this paper we know, designing an instructional online agent system should consider learner and system. More focus on learner at the beginning would be reduced failure factors. Defining the learners by intelligence test can probe their personal Learning style. It can be classified learner as defined thinker in suitable groups. Our intelligent agent should achieve two goals that are to
reach four responsibilities and controlled by system. Finally as the result of survey shows that relative different intelligences exist between students and department. In the future, with this paper we will define algorithms and term conditions for system and intelligent agent to determine actions for classifying learners. We will define interactive rules between system, intelligent agent and learners within an instructional environment.
REFERENCES [1] Baylor, A. L., & Jafari, A., “Intelligent agents for education:” Association for Educational Communication and Technology, Long Beach, 2000. [2] AMY L. Baylor “Agent-base Learning Environments as research tool for investigating teaching and learning”, J. Educational Computing Research, Vol. 26(3) 227-248, 2002 [3] Mantyla, K., Gividen, R. “A step-by-step Guide for Trainers.”American Society for Training & Development 1997. [4] Payr, S. “Pedagogical agent in open & distance learning” Wien: Jubilaumsfond of Nationalbank. Project No.8631, 2001. [5] Dreyfus, H.L., and Dreyfus, S. E., “the challenge of merleau-ponty’s phenomenology of embodiment for cognitive science.”, In Perspectives on embodiment, ed. G. Weiss and H.F. Haber, 103-119, 1999. [6] Sternberg, R.J. and Grigorenko,E.”Intelligence, Heredity and Environment.” New York:Cambridge University Press 1997. [7] Neisser, U., Boodoo, G., Bouchard, T.J., Boykin, W.A., Brody, N.,Ceci, S., Halpern, D.F., Loehlin, J.C., Perloff, R., Sternberg, R.J. & Urbina, S. ”Intelligence: Knowns and unknowns.” American Psychologist, 51, 2 77-101. [8] Howard, G. “Multiple intelligences: the theory in practice / Howard Gardner”1993. [9] Campbell, L., Campbell, B. & Dickinson, D. ”Teaching and learning through Multiple intelligence”, U.F.A university 1999 [10] O’connor, J. “NLP workbook a practical guide to achieving the results you want”, Martins The Printers Limited, Berwick upon Tweed Thorsons 2001. [11] Deporter, B. with hernacki, M. “Quantum Learning unleash the genius within you”, Judy Piatkus(Publishers)Ltd 1993. [12] Baylor, A.L. “Intelligent agents for education.” Paper presented at the American Educational Research Association, Montreal, Canada 1999.