Case-Based Reasoning (CBR) and Ability Development of Problem Solving

Yana Aditia Gerhana1, As’ari Djohar 2, Ayu Puji Rahayu3

1, 2.   SPS UPI Bandung

  1. STKIP Garut

 Jl. Dr. Setiabudi No. 229, Bandung, Jawa Barat, Indonesia

yanagerhana@gmail.com

 Abstract

This paper provides a conceptual overview of how artificial intelligence can be used in learning process. Case-based Reasoning (CBR) is part of artificial intelligence that provides a model of learning problem-solving. Problemsolving in the Case-based Reasoning is done by reusing previous solutions that have similarity. Thedevelopment of Case-based Reasoning is strongly influenced by cognitive science, many studies have proven the success of the Case-based Reasoning in learning. Case-based Reasoning is able to be an alternative solution to develop problem-solving skills of students in learning.

Keywords: Artificial Intelligence, Case-based Reasoning, Problem Solving,Similarities,Cognitive

 Introduction 

Gagne (Dahar 2011:119) states that the highest order of intellectual skills is a problem solving ability. The low ability of a teacher to develop problem solving skills of students is regarded to be one factor contributing towards the low competence of students. In performing learning activity there is a tendency for teachers to develop learning materials more by providing as many materials as possible, in hope that students will be able to understand and apply the knowledge acquired, rather than developing reasoning ability of students in solving problems .

The application of Case-based reasoning (CBR) in learning activity can be an alternative solution in developing problem solving skills for students. CBR is an automated reasoning system, where the problem is solved by means of utilizing past experience as revealed by Mulyana and Hartati (2009:17) that the CBR is a major paradigm in automated reasoning and machine learning, students who perform reasoning can solve new problems by seeing their similarities to one or several of previous problem solvings. Scank and Kolodner (Mulyana and Hartati 2009:19) reveal many studies have described the role of CBR in reasoning and learning for highly developed people. Kolodner (Mulyana and Hartati 2009:19) mentions an example of teaching formed by the CBR in reasoning diagnosis in the medical field of which one of the main components has used the type of pattern matching, which the process of case-based reasoning is based on the experience of previous patients .

  1. ArtificialIntelligence

Simon (Kusrini 2006:3) revealed that artificial intelligence is a research field, application and instruction related to computerprogramming to do smart things in the view of human being. More specific definition is proposed by Russell and Norving (2010:2), artificial intelligence is defined into four categories, one of which defines artificial intelligence from the perspective of thinking humanly, the definition is proposed by Bellman (Russell and Norving 2010:2) that artificial intelligence is the automation of activities, the activities in which we connect with the thinking humanly, activities consist of the decision-making, problem solving and learning. Furthermore, Russell and Norving (2010:3) explain the artificial intelligence from the perspective of human thinking in the cognitive approach, cognitive disciplines that unify the computer models of artificial intelligence as well as experimental techniques from cognitive psychology to try to build appropriate theories to examine the ways of human mind workings.

Intelligence was created and put into a machine (computer) in order that it can do the job as humans do. At the beginning of creation, the computer is only used for calculating. But along with age development, the role of computer is present in almost all aspects of human life. Computer is no longer used as a tool of counting. Moreover, it is expected that computer be empowered to do all the things that humans do.

2.1 The concept of Artificial Intelligence

        Kusrini (2006:5) states that there are some concepts in artificial intelligence, as follows:

  1. Turing Test – Methods of Testing Intelligence

Test is a method of testing intelligence created by Alan Turing. The testing process involves a questioner (human) and two objects are questioned. The one is a man and the other is a machine to be tested. Questioner can not look directly at the object which is questioned. Questioner is asked to differentiatethe answers madeby computers and what by the human based on the answersof both objects. If the questioner can’tdistinguish the answers made by computers and what by the human Turing views the machine tested is regarded to be SMART.

  1. Symbolic processing

Computer was originally designed to process numbers or figures (numerical processing). While the human tend to besymbolic in thinking and solving problem, not based on some formulas or doing math computation. The important charasteristic of artificial intelligence is that artificial intelligence is a part of computer science that make symbolic processing and non-algorithmic problem solving .

  1. Heuristic

The term heuristic is taken from the Greek word that means ‘find’. Heuristic is a strategy to make the search processing of problem space selectively, which guides the search process that we do along with the path that has the greatest chance of success.

  1. Conclusion(inferencing)

Artificial intelligence tries to make machine have the ability to think or consider (reasoning). Thinking ability (reasoning)including processes of conclusions (interferencing) is based on the facts and rules using heuristics or other search methods.

  1. Pattern Matching

Artificial intelligence works with pattern matching method that attempts to explain the objects, events or processes, in computational or logic relationships.

Artificial intelligence has been the basis of the development of CBR;this is in line with the statement of Hullermeier (2007:30) that the CBR is one of the latest developments in artificial intelligence research that has become technology. In CBR knowledge is stored in computer systems into a knowledge base that can be used in solving problem. The system was developed in order to have the thinking ability (reasoning) to reach the conclusion of the problem heuristically. CBR is a problem-solving model by matching the similar case of the previous problem.

  1. Case – Based Reasoning (CBR)

          The definition of CBR is expressed by Montani and Jain (2010:8) that CBR is a problem solving method that gives priority to past experience utilizing to solve current problems, the solution to the current problem can be found by reusing or adopting a solution to a problem that has been solved previously. Aamodt and Plaza (1994:2) state that CBR is basically used to solve a new problem by remembering a situation / same previous problem and using information and situation to solve the problems. In one illustration Aamodt and Plaza (1994:2) saidCBR is illustrated as problem solving situation by a Doctorwhen diagnosing one of his patients, the doctor remembered another patient whom he treated some times ago. The doctor thought the patient to another because of the similarity of symptoms disease patients. The doctor then uses data from earlier diagnosis and treatment of patients to determine the diagnosis and treatment of other patients. In addition, Aamodt and Plaza (1994:7) explain problem solving cycle in CBR system, which is described in Figure 1.

  siklus cbr Figure 1. CBR Cycle

(Aamodt and Plaza 1994:8)

 1. Retrieve

Regaining cases most relevant (similar) to a new case. The retrieval phase begins by describing / outlining some of the problems, and ends if there is a matching to the previous problem that has the highest level of compatibility. This section refers to the terms of identification, initial match, search, selection and execution.

  1. Reuse

Modeling / reusing old knowledge and information based on the most relevant quality of similarity into a new case, so as to produce the proposed solution where an adaptation to the new problem may be required.

  1. Revise

Reviewing a proposed solution and then testing it on a real case (simulation) and if necessary improving the solution to match the new case.

 4.  retain

Integrating / keeping new cases that have had a solution that can be used by following cases that are similar to the cases. However, if the new solution gets failure, improving and testing the used solutions tell the failure.

3.1 CBR and Human Reasoning

          Humans are creatures that have ability to think, so that the nature of humanis that they are thinking creature. Suriasumantri (2007: 42) states that reasoning is a process of thinking in drawing conclusions in the form of knowledge. The equation of CBR and human reasoning, Pal and Shiu (2004:5) argue that:

The process in CBR is like reasoning reflections of human. When confronted in a situation, the problem is solved by the human like the completion of the CBR. When facing a new problem it will refer to the same problems in previous days, both referring to the experience themselves or other people’s experiences are stored in memory.

Just as humans are capable of reasoning, CBR is developed to perform reasoning like human beings, through reasoning, CBR can do the matching and recalling solutions stored in previous days used to solve the current problems.

3.2 CBR in Education

          Kolodner et al (2003:3) stated thatlearning in the CBR paradigm means expanding one’s knowledge throughincluding new experiences into memory / database to be used in solving problems in the future time. Ritcher and Aamodt (2006:1) said that the development of CBR is greatly influenced by the results of the research field of cognitive science. Mulyana and Hartati (2009:19) stated that CBR is currently based on research on the role of memory in knowledge, Memory Organizing Packets (MOPS) has function to control the sequence of events, MOPs sets single event called “memory” and this memory is playing many roles of interpretation and problem solving.

1. Learning Theory Associated with CBR

4.1 Cognitive Learning

          The theory of cognitive learning is based on a view of Leibnitz concerning human nature. According to Leibnitz (Sanjaya 2010:236):

A human is active organism. Human constitutessource of all activities. As a matter of fact, human is free to act; human is free to make a choice in every situation. Freedom is the central point of his own conscience.

          A cognitive point of view of learning was revealed by Woolfolk (2009 : 4) as a general approach that sees learning as an active mental process of acquiring, remembering, and using knowledge .

          There are some studies on CBR (Kolodner, 2002; Richter and Aamodt, 2006; Pal and Shu, 2004; Lenz et al. 1998; Schank and Abelson, 1977) they agreed that the CBR is strongly influenced by cognitive science. CBR research is strongly influenced by the study of human history knowledge, especially about the role of human memory in knowledge. Human memory has a role in interpretation and problem solving.

4.2 Constructivistic Learning Theory

          Constructivistic learning theory of Sanjaya (2010:237) is included in one of cognitive categories. Joyce et al (2009:13) expressed the concept of learning, thatlearning is knowledge construction. Furthermore, in the process of learning, brain stores information, processes it, and change the previous conceptions. Learning is not just a process of absorbing information, ideas and skills, because these new materials will be constructed by brain.

According to Bruning et al (Woolfolk 2008:145) that students are active in constucting their own knowledge and social interaction is important in knowledge construction. In line with Bruning, Woolfolk (2008:145) says constructivism sees learning more than just receiving and processing information conveyed by the teacher or text.

Problems solving needs the ability to construct knowledge. The construction of knowledge in CBR system is provided through the medium of interaction, where students find similarities ofprevious problems solving in the past and adopting themto solve new problems.

4.3 Learning Theory of Inductive Thinking

Learning to think inductively that is initially pioneered by Hilda Taba (Joys et al, 1992:116) is designed to improve thinking ability. Joys et al (2009:100) reveals that students are natural conceptors who always perform conceptualization any time, compare and contrast all objects, events and emotions. In line with this natural tendency, it is imperative to form an effective learning environment that can lead students to improve their effectiveness in shaping and using conceptual skills in problem solving.

Just as inductive learning,the core ofCBR learning is emphasizing on developing the thinking ability of students to solve problems. The process of retrieving CBR is the first step to be followed when finding new problems. Retrieving process will perform two processing steps, namely the recognition of issues or facts and finding similarities of problem or the facts on the database to find similarities and conclutions.

4.4 Problem-Based Learning (PBL)

          Jonassen (2004:21) states that learning to solve problem constitutes the most important skills from which students can learn in any setting.

Hmelo-Silver et al (Eggen and Kauchak 2012:307) defines problem-based learning as a set of model that uses problem as a focus for developing problem-solving skills, subject and self-regulation. Furthermore, Hmelo-Silver et al, explains the characteristics of problem-based learning in Figure 2.

characteristik of PBLFigure 2. Characteristics of Problem-Based Learning

(Silver et al 2004)

           First of all, learning begins with a problem and problem solvingis the purpose of learning. Krajcik and Blumenfeld (Eggen and Kauchak (2012:307) said that problem-based learning activity is originated from a problem and problem solvingis the focus of learning. Secondly, students are responsible for developing strategies and solving the problem. Problem-based learning is implemented in small groups, so that all students are engaged in the problem solving process. Thirdly, teachers guide students through giving questions and providing support for other learning when students try to solve the problem.

In encounteringlearning issues, Schwartz et al (Eggen and Kauchak 2012:322) reveals that experts have tried to use technology to present the complex problems of real world. The same thing is expressed by Krajcik and Blumenfeld (Eggen and Kauchak 2012:323), the software designers have developed simulation of problem solving. Furthermore, Triona and Klahr (Eggen and Kauchak 2012:323) assertsa number of studies show that simulation produces as good learning as direct experience with concrete materials.

Kolodner et al (2003:2) says that problem-based learning(PBL) is parallel with case-based reasoning (CBR). CBR methodology provides classroom learning activity in which learning situations is in problem solving, while PBL provides reflections on the central role of problem-solving activities and determining roles for student as researcherwho discovers knowledge and the teacher as a facilitator.This is a constructivistic process. Furthermore, Kolodner et al (2003:2) says that the PBL and CBR are two approaches that complement each other and provide a solid foundation in constructivistic learning practices, or in other words PBL facilitates CBR to put philosophy into learning practice. Learning practice using CBR in framework of PBLuses the aid of information technology.

  1. Conclusion

          The change of learning paradigm has brought shift from teacher-centered learning to student-centered learning, which is characterized by a critical, creative and innovative attitude in solving problems. As automated reasoning and machine learning, CBR has been widely developed and adapted in many different fields, one of which is education. Through CBR students can develop problem solving skills with the help of information technology, which in turn will improve the quality of learning CBR itself.

References

Aamodt, A. dan Plaza E. (1994). “Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches”. Journal of Case-Based Reasoning: Foundational Issues, Methodological Variations, and System Approaches. AI Communications. IOS Press, Vol. 7: 1, pp. 39-59.

Dahar W. R. (2011). Teori-teori Belajar dan Pembelajaran. Bandung: Erlangga.

Eggen, P. dan Kauchak, D. (2012). Strategi dan Model Pembelajaran: Mengajarkan Konten dan Keterampilan Berpikir (edisi keenam). Boston: Pearson Education, Inc.

Joyce, B. Weil, M dan Calhoun, E. (2009). Models Of Teaching: Model-Model Pengajaran (edisi kedelapan). Boston: Pearson Education, Inc.

Joyce, B. Weil, M dan Calhoun, E. (1992). Models Of Teaching (forth ed.). Massachusetts: A Divition of Simon & Chuster, Inc.

Jonassen D, H. (2011). Learning to Solve Problems: A Handbook for Designing Problem-Solving Learning Environments. New York: Taylor & Francis Group.

Jonassen D, H. (2004). Learning to Solve Problems: An Intructional Guide. New Jersey: John Wiley & Sons, Inc.

Kolodner, J. L. (2002). “Analogical and Case-Based Reasoning: Their Implications for Education”. The Journal of The Learning Sciences, 11(1), 123–126,Lawrence Erlbaum Associates, Inc.

Kolodner, J. L, Hmelo, C. E, dan Narayan, N. H, (2003). Problem-Based Learning Meets Case-Based Reasoning. [Online]. Tersedia: http://www.cc.gatech.edu/projects/lbd/pdfs/pblcbr.pdf

Kolodner, J. L. Cox, M. T. dan Gonzálezcalero, P. D. (2005). “Case-based reasoning-inspired approaches to education”. Journal of The Knowledge Engineering Review, Vol. 00:0, 1–4. Cambridge University Press.

Lenz, et al. (1991) Case-BasedReasoning Technology From Foundations to Applications. Berlin: Springer.

Montani, S. dan JainL, L. C. (2010). Su ccessf ul Case-Base d Re as oning Appl icat ions – 1. Berlin: Springer.

Mulyana, S. dan Hartati, S. (2009). Tinjauan Singkat Perkembangan Case–Based Reasoning. Jurnal Seminas IF. ISSN: 1979-2328.  

Munir. (2009). Pembelajatan Jarak Jauh Berbasis TIK. Bandung. Alfabeta

Pal, S. K dan. Shiu, C.K S. (2004). Foundation of Soft Case-Based Reasoning. New Jersey: John Wiley & Sons, Inc.

Suriasumantri, S. J. (2009). Filsafat Ilmu: Sebuah Pengantar Populer. Jakarta: PT. Penebar Swadaya.

Sanjaya, W. (2010). Kurikulum dan Pembelajaran: Teori dan Praktik KTSP. Jakarta: Kencana Prenada Media Group.

Sanjaya, W. (2012). Perencanaan dan Desain Sistem Pembelajaran. Jakarta: Kencana Prenada Media Group.

Woolfolk, A. (2009). Educational Psychology: Active Learning Edition (ten ed). Boston: Pearson Education, Inc.

Iklan

Tinggalkan Balasan

Isikan data di bawah atau klik salah satu ikon untuk log in:

Logo WordPress.com

You are commenting using your WordPress.com account. Logout / Ubah )

Gambar Twitter

You are commenting using your Twitter account. Logout / Ubah )

Foto Facebook

You are commenting using your Facebook account. Logout / Ubah )

Foto Google+

You are commenting using your Google+ account. Logout / Ubah )

Connecting to %s