Designing an Assistance Tool to Support the Tutor in the Settingup of Learning Situations

1
Designing an Assistance Tool to Support the
Tutorin the Setting-up of Learning Situations
Interaction Collaborative
Téléformation Téléactivités
Elise Garrot, INSA de Lyon

• Meeting e-tutor
• 3rd February, 2006

2
Introduction

• Research area Computer-Based Environments for
  Human Learning.
• My work Identification and instrumentation of
  tutors activity in collaborative learning
  situations.
• Aims
• To model on-line tutoring centred on the tutor.
• To develop and implement a tool to assist the
  tutor in the setting-up of learning situations.

Partnership
3
Summary

• Research issues
• Modelling of on-line tutoring
• Design of the assistance tool
• Implementation of rules to advice the tutor
• Conclusion and futures directions

4
Research issues
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion
Authoring tools
Learners
Assist to design
Author
Carry-out activities
Carry-out activities
Extract activities traces
Design
Design
Generic Learning situations
Instantiated Learning situations
Learning situations
Monitoring tools
Interact
Are modelled in
Adaptsituations
Assist to monitor
Assistance tool to set-up learning situations
Tutor
Assist to set-up
Setting-up
Learning session
5
Summary

• Research issues
• Modelling of on-line tutoring
• Design of the assistance tool
• Implementation of rules to advice the tutor
• Conclusion and futures directions

6
Modelling of on-line tutoring
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Aim to identify tutors activity in
  collaborative learning situations.
• Fourteen diagrams developing
• Actors of the course tutor, learners, author
• Learners characteristics being able to influence
  their learning.
• Pedagogical and technical resources.
• Educational variables roles and degrees of
  freedom given to tutors.
• With the software MOT (Modeling using Object
  Types), an object-oriented modeling tool,
  developed by the Tele-University of Quebec
  (Paquette, 1996).

7
Modelling of on-line tutoring example
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion
8
Summary

• Research issues
• Modelling of on-line tutoring
• Design of the assistance tool
• Implementation of rules to advice the tutor
• Conclusion and futures directions

9
Development choices
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• To separate knowledge and reasoning
• Ontology Inference engine
• Advantages of ontology
• To make explicit what is regarded as implicit in
  the field.
• To use a vocabulary comprehensible by all actors.
  
• To associate properties to each concept.

Tutoring model MOT
Inference rules Jess
Ontology Protege2000
Plugin JessTab
10
Ontological model Actors (classes and slots)
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Three main concepts Actors, Actors-characteristic
  , learning-group-characteristics.
• These characteristics are constant during a
  course session.

11
Ontological model Actors (forms and instances)
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion
12
Ontological model Learning situations
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Structure of the course
• Course -gt Learning situation -gt Activity -gt
  Resource -gt Media

13
Ontological model Chronological record of
learning
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Data indicating learners achievements,
  interactions and carried-out activities.
• Tutors perceptions about learners at some point
  of the course level of initiative, motivation,
  stress, autonomy.
• Data concerning carried-out activities by a
  learning group.

14
Summary

• Research issues
• Modelling of on-line tutoring
• Design of the assistance tool
• Implementation of rules to advice the tutor
• Conclusion and futures directions

15
Implementation of rules to advice the tutor
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• An inference engine Jess infer on ontology
  instances to advice the tutor.
• Different types of inference rules
• Rules which create links between learners
  characteristics.
• Rules which deduct pedagogical advice to give to
  the tutor, according to some learners
  characteristics.
• Rules which create links between activities
  parameters and learners characteristics to
  advice the tutor an activity to attribute to a
  learner or a learning group.

16
Implementation of rules to advice the tutor
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Example to advise the tutor if a learner is in
  trouble.
• If all members of the learning group are also in
  trouble
• Help the group guide, provide ressources.
• If the learner is the only one to be in trouble
• If the learner usually does not interact with the
  group (cf. chronological record)
• Incite him/her to take part in the activities
  and to interact with the group.
• If the learner usually interact with the group
• Help the learner to overcome the difficulty by
  inciting him/her to have more reflection,
  suggesting errors, providing direct assistance
  (by giving a ressource or a part of the solution).

17
Summary

• Research issues
• Modelling of on-line tutoring
• Design of the assistance tool
• Implementation of rules to advice the tutor
• Conclusion and futures directions

18
Conclusion
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• Modelling of on-line tutoring (14 diagrams).
• Implementation of an expert system to assist
  tutors in setting-up and adapting learning
  sessions to learners and learning groups
• Based on an ontology learning concepts (actors,
  actors characteristics, learning situations and
  activities parameters, resources), their
  properties and relations and a chronological
  record.
• Inference engine rules which apply on the
  instances of the ontology to deduce pedagogical
  advice.

Feasibility of the system and many prospects
offered.
19
Futures directions
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion

• We have in prospect
• To develop precise educational rules,
• To develop a tutor interface to give him/her the
  possibility to modify and create learning rules,
• To associate a monitoring agent to automatically
  feed the system with data concerning activities.

20
Futures directions
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion
21
Futures directions
Research issues Tutoring model Design of the
tool - Implementation of rules - Conclusion
Authoring tools
Learners
Assist to design
Author
Carry-out activities
Design
Extract activities traces
Generic Learning situations
Instantiated Learning situations
Monitoring agent
Interact
Are modelled in
Adaptsituations
Assist to monitor
Assistance tool to set-up learning situations
Tutor
Assist to set-up
Provide data during the activities
Setting-up
Learning session