Presentation is loading. Please wait.

Presentation is loading. Please wait.

AN ADAPTIVE E-LEARNING SYSTEM BASED ON USERS' LEARNING STYLES

Similar presentations


Presentation on theme: "AN ADAPTIVE E-LEARNING SYSTEM BASED ON USERS' LEARNING STYLES"— Presentation transcript:

1 AN ADAPTIVE E-LEARNING SYSTEM BASED ON USERS' LEARNING STYLES
Author: Phạm Quang Dũng

2 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion

3 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 3

4 Introduction Motivation and problem statement
Each learner has his own individual needs and characteristics Most of LMSs do not consider learners’ needs and preferences  the need for providing learners with adaptive courses While adaptive systems support adaptivity, they support only few functions of web-enhanced education, and the content of courses is not available for reuse. In contrast, LMSs focus on supporting teachers and help to make online teaching as easy as possible.  use an adaptive learning management system

5 Introduction Research issues
1. How can learning styles be identified? Find a literature-based method for automatic identifying learners’ learning styles based on their behaviour and actions on learning objects in online courses using LMSs suitable for LMSs in general 2. How can adaptive courses be provided in LMSs? which types of learning objects their order

6 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 6

7 Learning object? any digital resource that can be reused to support learning (D.A. Wiley, 2000) digital images or photos, video or audio snippets, small bits of text, animations, a web page Characterstics Share and reuse Digital Metadata-tagged Description information: title, author, format, content description, instructional function Instructional and Target-Oriented

8 Learning style models To classify and characterise how students receive and process information. Refer to fundamental aspects: cognitive style learning strategy Well-known models: Myers-Briggs, Kolb, Felder-Silverman

9 Learning style models Felder–Silverman Learning Style Model
Each learner has a preference on each of the four dimensions: Active – Reflective learning by doing – learning by thinking group work – work alone Sensing – Intuitive concrete material – abstract material more practical – more innovative and creative patient / not patient with details standard procedures – challenges Visual – Verbal learning from pictures – learning from words Sequential – Global learn in linear steps – learn in large leaps good in using partial knowledge – need “big picture”

10 Learning style models - FSLSM (cont’) Types of combination of LS dimensions
active/sensing/visual/sequential active/sensing/visual/global active/sensing/verbal/sequential active/sensing/verbal/global active/intuitive/visual/sequential active/intuitive/visual/global active/intuitive/verbal/sequential active/intuitive/verbal/global reflective/sensing/visual/sequential reflective/sensing/visual/global reflective/sensing/verbal/sequential reflective/sensing/verbal/global reflective/intuitive/visual/sequential reflective/intuitive/visual/global reflective/intuitive/verbal/sequential reflective/intuitive/verbal/global 10

11 FSLSM (cont’) Index of Learning Style (ILS) questionnaire
44 questions, 11 for each LS dimensions Scales of the dimensions: 11

12 A reductive questionnaire
Based on FSLSM To be used for collecting initial learning style information of students Aims at saving time for students to answer Contains of 20 questions some from the ILS questionnaire, the rest from us 5 questions for each LS dimension The questionnaire Graphical presentation:

13 Implications of LSs in education
make learners aware of their learning styles and show them their individual strengths and weaknesses students can be supported by matching the teaching style with their learning styles

14 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 14

15 Ontology in education Ontology represents domain knowledge by defining terminology, concepts, relations, and hierarchies Ex. of educational ontology: OntoEdu It enables education applications to share and reuse educational content Ontology is machine-readable and reasonable: Suitable for description of learning objects It will be faster and more convenient to query and retrieval educational material

16 Intelligent agents in education
how to provide adaptive teaching which is suitable to each student? the use of Artificial Intelligence (AI) techniques such as Multi Agents or Agent Society-based architectures intelligence may be applied through user models to make assumptions about the user’s state of knowledge, which may in turn help determine the user’s learning needs may enable the system to dynamically personalise applications and services to meet user preferences, goals and desires

17 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 17

18 Introduction to LMSs Developed for teachers to create and manage their courses. Can be built based on pedagogical strategies: more learner-centered or more teacher-centered The applied strategies focus mainly on how to teach learners from a general point of view, without considering the individual needs of learners.

19 Adaptivity in LMSs Adaptivity indicates all kinds of automatic adaptation to individual learners’ needs. Course’s content Personal annotations

20 Benefits from using the Felder-Silverman learning style model in LMSs
FSLSM describes learning style in more detail, represents also balanced preferences  allows providing more accurate adaptivity FSLSM considers learning styles as “flexibly stable”  LSs might change over time. An adaptive system can adjust to the change. FSLSM considers learning styles as tendencies  a student might act differently from his LS tendency. An adaptive system should consider also exceptions and extraordinary situations.

21 Behaviour of learners in LMSs with respect to learning styles
Active/Reflective dimension Active learners: do exercise first then look at examples perform more self-assessment questions Reflective learners: visit examples first then perform exercises spend more time on examples and outlines performed better on questions about interpreting predefined solutions

22 Behaviour of learners Benefits
Make teachers and course developers aware of the different needs, different ways of learning of their students. Should provide courses with many different learning materials that support different learning styles. Might present learning materials in different orders corresponding to different preference for LSs.

23 Providing adaptive courses in LMSs
Course elements Adaptation features 23

24 Providing adaptive courses in LMSs Course elements
A course consists of several chapters, where for each chapter, adaptivity can be provided. Each chapter includes: An outline Content objects definitions, algorithms, graphics, etc. Examples Self-assessment tests Exercises A summary

25 Providing adaptive courses in LMSs Adaptation features
Indicate how a course can change for students with different learning styles. Include: the sequence of LOs and their positions. the number of presented examples and exercises

26 Adaptation features (cont’)
For active learners: outlines are only presented once before the content objects the number of exercises is increased a small number of examples is presented self-assessment tests are presented at the beginning and at the end of a chapter a final summary is provided in order to conclude the chapter

27 Adaptation features (cont’)
For reflective learners: the number of exercises and self-assessment tests is decreased content objects are presented before examples outlines are additionally provided between the topics a conclusion is presented straight after all content objects

28 Methodology of incorporating LSs in a LMS
Creating adaptive course Course structure Learning objects with learning style properties enough interchangeable LO? Student modelling A LS questionnaire for initialising An automatic approach for revising Providing adaptive course Combination of selecting and ordering learning objects

29 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 29

30 Problems with collaborative student modelling that use a questionnaire
Uncertainty because of: a lack of students’ motivation a lack of self-awareness about their learning preferences the influence of expectations from others Questionnaires are static and describe the learning style of a student at a specific point of time The result depends much on students’ mood

31 Benefits of using automatic student modelling
does not require additional effort from students is free of uncertainty can be more fault-tolerant due to information gathering over a longer period of time can recognise and update the change of students’ learning preferences

32 Automatic student modelling approaches

33 Automatic student modelling approaches data-driven vs. literature-based

34 Automatic student modelling The data-driven approach
uses sample data in order to build a model for identifying learning styles from the behaviour of learners aims at building a model that imitates the ILS questionnaire Advantage: the model can be very accurate due to the use of real data Disadvantage: the approach strictly depends on the available data and is developed for specific systems

35 Automatic student modelling The literature-based approach
uses the behaviour of students in order to get hints about their learning style preferences then applies a rule-based method to calculate LSs from the number of matching hints Advantage: generic and applicable for data gathered from any course Disadvantage: might have problems in estimating the importance of the different hints

36 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 36

37 Methodology for implementing adaptation
Annotating learning objects Estimating learning styles Providing adaptivity 37

38 Methodology Annotating learning objects
Each learning object is annotated with one subtype of any element in the set of 16 types of combination E.g: Annotation of an example LO is RefSen Active Reflective Sensing Intuitive Visual Verbal Sequential Global Self-assessment exercises, multiple-question- guessing exercises Examples, outlines, summaries, result pages Examples, explanation, facts, practical material Definitions, algorithms Images, graphics, charts, animations, videos Text, audio Step-by-step exercises, constrict link pages Outlines, summaries, all-link pages 38 38

39 Methodology Estimating learning styles
Expected time spent on each learning object, Timeexpected_stay, is determined. The time that a learner actually spent on each learning object, Timespent, is recorded. Ratios for number of visits with respect to each LS element 39

40 Methodology Estimating learning styles (cont’)
An example  Learning style: moderate Active/Reflective, and strong Visual. Ravg LS Preference 0 – 0.3 Weak 0.3 – 0.7 Moderate 0.7 – 1 Strong ACT REF SNS INT VIS VRB SEQ GLO Ravg 0.5 0.6 0.25 0.2 0.8 0.15 0.9 40

41 Methodology Providing adaptivity
Assumption: interchangeable learning objects are sufficient for each learning content. The LMS automatically delivers suitable LOs for each learner based on: What learning content he choses His learning style that has been identified Previous example: only LOs with Act/Ref/Vis annotations. Combined with changing their appearance order 41

42 System’s adaptation

43 System’s domain ontology

44 Outline Introduction Learning objects Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 44

45 System architecture A multi-agent one with artificial agents 45

46 System interface and functionality
Administrator: updates personal information of teachers and students, views statistics about each individual or all of students' behaviour with respect to FSLSM other management tasks

47

48 System interface and functionality Teachers
update list of his courses: subjects, chapters, sections update his learning objects: outlines, definitions, algorithms, graphics, examples, exercises, summaries, etc. set up tests and see participated students' results accept application requests for his course from students view statistics of students' behaviour related to their learning styles

49

50 System interface and functionality Students
register for a course take registered courses do the tests see the test results

51 System interface and functionality System’s agents
Learning style monitoring agent keeps track on every student's number of and his visit spent time on learning objects of the courses stores students' learning styles and updates new estimated ones Adaptive content agent chooses and orders the learning objects to present for each student

52 LS detection result Experiment:
an Artificial Intelligence course – 9 weeks 204 learning objects – test of LS properties 44 participated students – were asked to fill in the Index of Learning Style (ILS) questionnaire Precision: (72,73%, 70.15%, 79.54%, 65.91%) for Act/Ref, Sen/Int, Vis/Vrb, and Seq/Glo 52

53 Outline Introduction Learning objects and Learning styles
Ontologies and intelligent agents in education Incorporation of learning styles in a learning management system Automatic detection of learning styles in LMSs Conception of an adaptive e-learning system POLCA – Implementation and results Conclusion 53

54 Contributions Develop a reductive questionnaire for detecting learning styles Make a survey of students' learning styles based on the Felder-Silver learning style model Develop an agent-based architecture for building adaptive LMSs in general Propose an annotation of learning objects and a mixture method to provide adaptivity in LMSs according to users' learning styles

55 Contributions Propose a new automatic and dynamic approach based on literature for identifying students’ learning styles in LMSs has a promising detection result, is simpler than existing ones, and can be applied for LMSs in general Develop an adaptive e-learning system incorporating above architecture and methodologies. 55

56 Limitations no incorporated communication channel among students
the short testing time and the restricted pools of testing students

57 Future work develop more system’s functions
have more accurate results in LS detection: include more students’ behaviour patterns examine more exceptions of student behaviour consider the ability of including the relationship between learning styles and cognitive skills focus on providing better adaptivity find whether there are adaptation features which have more impact than others monitoring agent will track also their learning performance

58 Thanks for your attention!

59 Summarise the most contributions - Section 10.1
Add the reasons why to use those appendices Add our own citations - Sections 8.1, 8.2, 9.1, 9.3.1 Explain more clearly about literature-based approach and Graf's method (including Figure 7.2 and Table 7.1) - Section 7.2 Make the comparison between our method with the others more clearly - Section 9.3.1


Download ppt "AN ADAPTIVE E-LEARNING SYSTEM BASED ON USERS' LEARNING STYLES"

Similar presentations


Ads by Google