Unformatted text preview:

[2] Bennet, R. E. et al. “Psychometric and Cognitive Functioning of an Underdetermined Computer-Based Response Type for Quantitative Reasoning”, http://ftp.ets.org/pub/gre/gre_95-11r.pdf, 1998.BiographyUSE OF MULTIMODAL INPUT INTEGRAGTION TO IMPROVE ANINTELLIGENT TUTORING SYSTEM FOR FIRST GRADE MATH STUDENTSJohn D. Donath, Adjunct Lecturer of Computer Science, Queens College, City University of NewYork, PHD Student, Graduate and University Center, City University of New York. Bethel Lynn, Adjunct Lecturer of Computer Science, Queens College, City University of NewYork, EdD Student, Columbia UniversityABSTRACT:Important elements ofeducation include theability to customize aprogram to a student'sneeds, and tocommunicate with thestudent on his/herlevel. A successfultutoring system is onethat can provide bothof these capabilities.Recent advances ineducationaltechnology andartificial intelligenceallow us to developsoftware that is able toassess a student'slevel, interests, andareas of difficulty inorder to present thestudent withinformation andproblems that willmaximize the student'spotential and ability tolearn. Developmentsin voice and inkrecognition, as well asadvances in systemsthat integrate the twoallow us to improvethe agents available tobetter communicatewith students,eliminating theoverhead associatedwith teaching anindividual how tointeract with acomputer. In thispaper, we present anoverview of differenttechnologies that areavailable forintelligent tutoringagents for first grademath students. Wealso present anoverview of Voice andInk recognitiontechnologies. Finally,we present our ideasand preliminaryresearch data on howa combination of thetwo technologies canhelp develop betterlearning systems ingeneral.Keywords: IntelligentAgents in Education,Multimodal LearningSystems.1. IntroductionArtificial Intelligencetechnology has allowedus to write intelligentsoftware that can learnbased on userinteractions and a set ofrules. ApplyingArtificial Intelligencetechniques toeducational programshas been the goal ofmuch research. This isdue to the power ofthese systems to adaptto the individual needsof the students,providing them with acustomized learningenvironment. Thisincludes the ability topresent topics in thebest possible order toenable the student tobuild upon his/herexisting knowledge, atthe student’s level.Also, review problemsshould be presented ina manner suitable forthe student. The goalof each question is tochallenge the student,but not to be toodifficult. AI techniquesare used to learn frompast habits and interactwith learners in orderto improve theireducational experience.Yet, one of thelimitations oftechnology is the formof communication withstudents. Students,especially young ones,benefit most fromsystems that cancommunicate with themin natural languages.Therefore, usingintelligent agents foreducation has theoverhead of needing toteach students how tointerface with thecomputer. Recentadvances in Voice andInk recognition yieldthe promise ofdeveloping agents formore naturalcommunication withstudents. In this paper, wepresent recent advancesin educationaltechnology andvoice/ink recognition,including ideas on howto improve them. Wethen present our ideasfor developing anintelligent tutoringsystem, which benefitsfrom the fusion of thesetechnologies.2. Current ResearchCurrently, variousagents exist that areutilized primarily indistance learningcourses. These agentsexist mostly on thecollege level. The firstof these is the DigitalTA, which monitors astudent’s progress. Ifthe student is late insubmitting assignmentsor quizzes, it sends thestudent a customized e-mail. It also checks astudent’s work againstother student’s work,looking for anypatterns and possiblecheating. Digital TAalso analyzes patternsin the work of thestudent in order todetermine the mostappropriate remindersand assign theproblems with themaximum benefit.Another technologythat exists is known asthe digital tutor. Adigital tutor drawsfrom information in adatabase about eachstudent. This wouldinclude the student’sprevious courses andgrades. As the studentadvances in a course,the system adjusts topresent the studentinformation andproblems on his/herlevel. In moreadvancedenvironments, a digitaltutor can serve as amedium ofcommunication amongstudents. [7]Student modelingevaluates the presentstatus of the student’sknowledge and tries toexecute usefulinstruction based on theevaluation. Studentmodeling associateswith the student’sapparent level ofcomprehension ofmaterials, then basedon the student’sknowledge. It willform deductions aboutthe student’s level ofcomprehension andhow they assimilatetheir new knowledge.[5]The student modelkeeps a record that isspecific to the student’sability on each topic.Its construction isbased on the student’sperformance onassignments includingquizzes and homework.The system then judgeshow well a topic asbeen acquired orlearned by the studentand this inference isthen used to establishwhich new subjectsshould be learned.[5]2.1 EducationalTheoryAbout 10 years ago,the Educational TestingService began offeringComputer AdaptiveTesting on some of itsstandardized exams.The idea behind themwas to present studentswith questions on theirlevel, allowing a moreaccurate gradeassessment with feweroverheads. Thealgorithm that theyused involveddeveloping questionson different levels. Thetest began with allstudents with receivingaverage level questions,and based on previousanswers, presenting thestudents withsubsequently easier ormore difficultquestions. While thismethod was a goodapproach for testing, amore generalizedapproach can be usedto develop intelligenttutoring


View Full Document
Download Study Notes
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Study Notes and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Study Notes 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?