Multi-user Virtual Environment to aid Collaborative Learning and Transfer of Scientific Knowledge and Inquiry Skills

minedesertSoftware and s/w Development

Oct 31, 2013 (4 years and 7 days ago)

71 views

Multi
-
user Virtual Environment to aid
Collaborative Learning and Transfer of
Scientific Knowledge and Inquiry Skills

Name: Nader Hanna

Student ID: 42259541

Supervisor: Prof. Deborah Richards

ITEC810 End of Semester Workshop
-

2012

Agenda


Introduction


What is Activity Theory?


The Aim of the Study


The Focus of the Study


Literature Review about on Using AT in Designing Collaborative
environment


Literature Review About the Approaches of Agent Collaboration.


Agent
-
to
-
agent Collaboration


One
-
way Human
-
agent Collaboration


Two
-
way Human Agent Collaboration


Proposed Framework of Multi
-
agent Collaborative Virtual Learning
(MACVILLE)


Proposed Collaborative Agent Architecture


Challenges


Conclusion

2

Introduction


Why collaborative learning?


Classroom learning improves significantly when students participate in
collaborative learning
[
Johnson and Johnson, 1999
]
.



Why virtual environment?


Using virtual environment increases the involvement of the students in learning
activity.



What is Intelligent agent?


A software agent that has some intelligence and autonomous in its environment , it
takes decisions by itself.



What is BDI agent reasoning model?


Belief
-
Desire
-
Intention model, agent has knowledge about his environment, plans
to achieve and methods to achieve these plans

3

What is Activity Theory?


Activity Theory (AT) is a theoretical framework for analyzing human practices in a
given context. It is used to arrange the collaborative learning.

1.
Subject: Members of the learning group

2.
Object: Experiences, Knowledge , Products,…

3.
Community: learning group, system, society ,…

4.
Tools:

plans, spoken language, ICT applications, devices

,…

5.
Rules:

practices, collaboration, negotiation

,…

6.
division of
labor:learners
, tutors, supervisors, virtual agents


Subject uses Tools to interact with Object.


Community uses Division of Effort to interact with


Object, and uses Rules to interact with Subject.

Subject

Object


Community

Rules

Division

of
Effort

Tools

Outcome

Structure of Activity theory

4

The aim of the study


The aim of this study is to design, implement, and test a collaborative
multi
-
agent multi
-
user virtual learning environment.


The collaborative environment will include 3 levels of collaboration:

1.
Human
-
human collaboration in physical world.

2.
Human
-
agent collaboration in virtual world.

3.
Human
-
human collaboration in virtual world.





5





The virtual environment

6

Levels of collaboration


0
5
10
15
20
25
30
1
2
3
4
WK1
WK2
WK3
Human
-
human physical
collaboration

Human
-
agent virtual
collaboration

1

2

7

Levels of collaboration
(cont.)


3

Human
-
human virtual
collaboration

8

The Focus of the Study


Collaborative
Learning

Learning
Scientific
Conclusion

Agent
Reasoning
Model

Activity Theory


We
target
this part

9

Literature Review on Using AT in Designing Collaborative
Environment


Miao [1] presents a conceptual framework for the design of virtual problem
-
based
learning environments in the light of activity theory.


Gifford and
Enyedy

[2] propose a framework called Activity Centered Design
(ACD). Their proposed framework is based on three main concepts of AT.


Liang

et

al
.

[
3
]

further

build

on

the

six

steps

to

define

components

and

their

relationships

for

collaborative

network

learning
.


Norris

and

Wong

[
4
]

use

AT

to

identify

any

difficulties

that

users

may

have

when

navigating

through

QuickTime

Virtual

Reality

Environments

(QTVR)
.


Zurita

and

Nussbaum

[
5
]

identified

six

steps

to

propose

a

conceptual

framework

for

mobile

Computer

Supported

Collaborative

Learning

(MCSCL)

activities
.


10

Literature Review on the approaches of agent
collaboration


We determined 3 possible categories of agent collaboration:

1.
Agent
-
agent Collaboration…interacting agents in a dynamic environment using
techniques such as
Tuple
-
Spaces, Group Computation and Roles.

2.
One
-
way Human
-
agent Collaboration…where agent behave as a mediator or
facilitator to human
-
human collaborative activities.

3.
Two
-
way Human
-
agent Collaboration…where agent directly interacts with
human actions.

11

Agent
-
to
-
agent collaboration


use of a shared/global model between agents to achieve the determined goal.


Agent Collaboration Approaches:


Tuple
-
spaces provide a multi
-
agent
-
like architecture, where agents can collaborate
through writing, reading or removing
tuples

in the space.


Group Computation a way to program group based activities .


Activity Theory is a framework to design the mediated interaction that may
happen between agents.


Roles are used to define common interactions between agents in virtual
environments.


12

One
-
way human
-
agent collaboration


actions performed by the human with assistance from the agent.


Aguilar et al. [6] Agent could assist the group during the execution stage of a
Team Training Strategy


Yacine

and
Tahar

[7] the role of the agent could be a mediator where the agent’s
role is to facilitate the collaboration between users and give feedback


Zhang et al. [8, 9]The role of the mediator agent could be to facilitate the
communication of a user with other users


Luin

et al. [10] the agent could collaborate with the user by answering the user’s
questions while navigating into a virtual world

13

Two
-
way human agent collaboration


the actions of both the human and agent interleave and depend on each other.


Miller et al [11] Humans and/or agents could be are working together to achieve a
goal


Lesh

et al. [12] plan recognition algorithm in order to reduce communication
during collaboration between a human and an agent


Miao et al. [13] Agent could behave and modify his
behaviour

according to human
actions, for example to train learners to handle abnormal situations while driving
cars


Hedfi

et al. [14] Collaboration between human and agent could take the form of
negotiation in design something


Fan and Yen [15] To make a teamwork between human and agent, agent may need
estimate its human partner’s cognitive load



14

Learner Companion

Learner Colleague

Agent

Collaborative
Virtual World

Learner
-
agent
group

Learner
-
learner
physical group

Learner
-
learner
online group


Interaction


Guidance


Discussion


Role

taking


Face
-
to
-
face
communication



Interaction


Guidance


Virtual

communication


Online

communication



Debating


Negotiating


Conclusion Making

Social Interaction

Collect evidences

Write report

3D character graphic

3D character animation


BDI


Co
-
Operation


Social


Agent reasoning model

Network Protocol

Database

Communication

Text

Awareness

Voice

Target Determining

Roles Specifying

Progression

Role playing

Virtual Collaborative

Collaboration

Discussion

Guidance

Decision
-
making

Roleplaying

KnowledgeSharing

NoteTaking

Conclusion Making

Decision Justifying

Idea Defending

Questioning

Comparing

Conclusion Making

Decision Updating

Negotiation

Proposed framework of
Multi
-
agent collaborative
virtual learning (MACVILLE)

Subject

Object


Community

Rules

Division

of
Eff ort

Tools

Outcome

15

Proposed collaborative
agent architecture



Situation

Goal Setting

Planning

Situation Understanding

Learning

Cognitive Processes

Plan

Executing/Adopting

Information Retrieval

Knowledge Integration

Knowledge/ Memory

Memory and its Processes

The Intelligent and Cognitive processes

Situation Understanding

Learner Realizing

Partner Progress Evaluating

Action and Communication

Social and Collaborative Processes

16

The Intelligent and Cognitive processes


The memory or the knowledge base is where
the agents store information, knowledge and
experience.


two processes related to the memory:
Knowledge Integration to add a new
experience to the stored knowledge, and
Information Retrieval to get the appropriate
piece of information



Information Retrieval

Knowledge Integration

Knowledge/ Memory

Memory and its Processes

17

The Intelligent and Cognitive processes (cont.)


Cognitive processes gives the agent
the ability to
perform the tasks
and change in its environment.


Situation
Understanding…using inputs from the
environment to understand where is the agent?


Goal
Setting…
using inputs from the
environment
and the human learner to determine what the
agent going to collaborate with the user.


Planning…the ability to have a different plans for
each task.


Executing/Adopting…adopt
the plans to perform
the task according to the changes in the
environment.


Learning…new rules may be added to knowledge
base for future usage.



Goal Setting

Planning

Situation Understanding

Learning

Cognitive Processes

Plan Executing/Adopting

18

Social and Collaborative Processes


Social and Collaborative
Processes…


Situation
Understanding…understanding what is
the social situation the agent going to handle.


Learner
Realizing…understanding who is the
human team member, and what is the role of
human in the collaboration activity.


Partner Progress
Evaluating…working in a
teamwork needs the members to feel the need to
others help.


Action and Communication
…instructions
exchanged between human and agent.




Situation Understanding

Learner Realizing

Partner Progress Evaluating

Action and Communication

Social and Collaborative Processes

19

Challenges


Conceptual and design challenges may include:

1.
Determining the factors that control human collaboration and determine
successful teamwork.

2.
The difficulties in capturing and measuring the levels and nature of collaboration
that take place between human and agent.


Implementation challenges may include:

1.
Creating, selecting or adapting an appropriate implementation framework to
extend the [BDI] reasoning model of the agent.

2.
Integrating the framework of agent reasoning with a 3D game engine already in
use (unity3D).


20

Conclusion


There are different approaches for agent collaboration in 3D virtual environment.


Activity Theory is a framework used to manage collaborative learning in physical
classroom, and could be use in virtual collaborative environment .


Two
-
way human
-
agent collaboration needs the agent to be fully aware of the
environment, the task to be achieved, the plan to be done and the partner.


BDI model of agent reasoning give the agent the ability to think in doing required
task regardless to participation from other actor.


Collaboration between human user and agent needs the agent to go beyond BDI
model.





21

References

1.
Miao, Y.: An Activity Theoretical Approach to a Virtual Problem Based Learning
Environment. In: the 2000 International Conference on Information in the 21
Century: Emerging Technologies and New Challenges, pp. 647
-
654. (2000)

2.
Gifford, B.R.,
Enyedy
, N.D.: Activity Centered Design: Towards a Theoretical
Framework for CSCL. In:
Roschelle
, C.H.J. (ed.) Proceedings of the 1999 conference
on Computer support for collaborative learning, pp. 22
-
37. International Society of
the Learning Sciences, Palo Alto, California (1999)

3.
Liang, X., Wang, R.,
Bai
, G.: A Multi
-
Agent System Based on Activity Theory for
Collaborative Network Learning. In: First International Workshop on Education
Technology and Computer Science (ETCS '09), pp. 392
-
397. (2009)

4.
Norris, B.E., Wong, B.L.W.: Activity Breakdowns in QuickTime Virtual Reality
Environments. Proceedings of the First Australasian User Interface Conference
(AUIC '00), pp. 67
-

72. IEEE Computer Society, Canberra, ACT (2000)

5.
Zurita
, G., Nussbaum, M.: A Conceptual Framework Based on Activity Theory for
Mobile CSCL. British Journal of Educational Technology 38,211
-
235 (2007)


22

References
(cont.)

6.
Aguilar, R.A., Antonio,
A.d., Imbert, R.: An Intelligent Collaborative Virtual Environment
for Team Training
--

A Preliminary Report. In: 15th International Conference on
Computing (CIC '06), pp. 236
-
239. (2006)


7.
Yacine
, L.,
Tahar
, B.: Supporting Collaboration in Agent
-
Based Collaborative Learning
System (SACA ). In: Information and Communication Technologies ( ICTTA '06 ), pp.
2843
-
2848. (2006)


8.
Zhang, C., Xi, J., Yang, X.: An Architecture for Intelligent Collaborative Systems Based on
Multi
-
agent In: 12th International Conference on Computer Supported Cooperative Work
in Design (CSCWD '08), pp. 367
-

372. (2008)


9.
Zhang, P., Li, X.: The Framework of Multi Intelligent Agent Based on Collaborative
Design. In: International Conference on Future
BioMedical

Information Engineering (
FBIE '09), pp. 513
-

517. (2009)


10.
Luin
,
J.v
.,
Akker
,
R.o.d
.,
Nijholt
, A.: A Dialogue Agent for Navigation Support in Virtual
Reality. In: extended abstracts on Conference on Human Factors in Computing Systems
(CHI '01), pp. 117
-
118. ACM, (2001)




23

References
(cont.)

11.
Miller, M.S., Yin, J.,
Volz
, R.A.,
Ioerger
, T.R., Yen, J.: Training Teams with Collaborative
Agents. In: Proceedings of the 5th International Conference on Intelligent Tutoring Systems,
pp. 63
-
72. Springer
-
Verlag
, 745826 (2000)


12.
Lesh
, N., Rich, C.,
Sidner
, C.L.: Using plan recognition in human
-
computer collaboration.
In: Proceedings of the seventh international conference on User modeling, pp. 23
-
32.
Springer
-
Verlag

New York, Inc., 317331 (1999)


13.
Miao, Y., Hoppe, U.,
Pinkwart
, N.: Naughty Agents Can Be Helpful: Training Drivers to
Handle
DangerousSituations

in Virtual Reality. In: Sixth International Conference on
Advanced Learning Technologies (ICALT '06), pp. 735
-
739. (2006)


14.
Hedfi
, R., Ito, T., Fujita, K.: Towards Collective Collaborative Design: An Implementation of
Agent
-
Mediated Collaborative 3D Products Design System. In: 2010 International
Symposium on Collaborative Technologies and Systems (CTS), pp. 314
-
321. (2010)


15.
Fan, X., Yen, J.: Realistic cognitive load modeling for enhancing shared mental models in
human
-
agent collaboration. In: Proceedings of the 6th international joint conference on
Autonomous agents and
multiagent

systems, pp. 1
-
8. ACM, 1329197 (2007)


24