Behaviorbased Cognitive Architecture for Meditative ELearning

1
Behavior-based Cognitive Architecture for
Meditative E-Learning
Hiran Ekanayake University of Colombo School of
Computing IITC 2006
2
Main Objective

• Integrate the techniques and knowledge from the
  modern cognitive science framework to the
  e-learning framework to automatically respond to
  learners dynamic psychological reactions to
  deliver an adaptive content, which will
  ultimately enrich the e-learning with pedagogical
  measures

3
Outline

• Conventional Learning
• The Paradigm Shift in Learning
• The Student-Centered Learning and E-Learning
• Cognitive Science Philosophy
• Psychological Aspects of Learning
• The Proposed Architecture
• Biofeedback Sensing Recovery of Emotions
• Arrangement of Modules
• Assumptions

4
Conventional Teacher-centered Learning
Teacher instructs. Students learn by following
instructions and observing postures.
5
Conventional Teacher-centered Learning

• Feedback
• Direct questions
• Indirect gestures, emotional mood

In the Learning process, continuous feedback is
an essential requirement to change the delivery
mechanism so that most students will get the
maximum benefit
6
The Paradigm Shift
Teacher-centered Learning
Student-centered Learning
7
Student-centered Learning

• E.g.
• E-Learning
• Features
• Low cost
• On-demand learning
• Distance learning

Indirect feedback is given little (no)
attention. It is difficult to adjust dynamically
the learning content according to learners
dynamic behavioral changes with respect to
psychology!
8
Traditional E-Learning Framework
The learner is observing a pre-arranged
presentation with limited interactions
9
Adaptive E-Learning Concept

• The sequence of the presentation is dynamically
  arranged based on,
• Learning patterns
• Motor actions (key presses, mouse movements, eye
  movements)

No attention for learners psychological
responses!
10
Making E-Learning a Pedagogy

• The Behavior-based Cognitive Architecture for
  Meditative E-Learning

11
Philosophy to Science
Greek philosophers, Plato and Aristotle, tried to
explain the nature of human knowledge. Study of
mind is a province of philosophy
In 19th century, experimental psychology
developed. Study of mind is now a province of
experimental psychology
Experimental psychology became dominated by
behaviorism, a view that virtually denied the
existence of mind
Primitive Computers Chunks as mental
representations Founding Artificial
Intelligence Rules as mental grammars
Advancements in brain scanning technologies, such
as, EEG, MRI, fMRI
Rebirth of Cognitive Science in mid 1990s as an
interdisciplinary study of mind and intelligence,
and emerging of the field Cognitive Modeling to
model descrete functions of human cognition as
computational models
12
Cognitive Sciences
The Cognitive Science is explained as an
interdisciplinary study of mind and intelligence,
where Psychology, Artificial Intelligence,
Linguistics, Neuroscience, Anthropology, and
Philosophy are studied as one.
Cognitive Modeling is computing that evolved to
simulate human problem solving and mental task
processes of a specific task in a computerized
model. These models are useful to simulate and
predict human behavior or performance of tasks
similar to the ones modeled Cognitive
architectures ACT-R, SOAR.
Affective Computing is computing that relates to,
arises from, or deliberately influences emotions.
Emotion is fundamental to human experience,
influencing cognition, perception, and everyday
tasks such as learning, communication, and even
rational decision-making.
13
Active Research Centered to Cognitive Science
United States Air Force Research Laboratory
Army Research Laboratory Carnegie Mellon
University George Washington University NASA
Ames Research Center United Kingdom Cardiff
University University of Sussex Canada,
Germany, Italy, Netherlands, France, Japan
14
The Learners Expected Behavior
External Stimuli
Psychological Reactions as Emotions
Biofeedback
Biological Response for Psychological Signals
Reflected in Autonomous Nervous System
15
The Proposed Architecture
Learner Profiles
Psychological Signals
Biofeedback Sensing of Psychological Signals
Cognitive Decision Making Module
Learning Content
Content Delivery Manager
The Conscious Interactions of the Learner
The Subconscious Psychological Reactions of the
Learner
16
Biofeedback Sensing Instruments
Source http//affect.media.mit.edu
17
Sensing of Biofeedback Signals

• Brain waves
• Heart rate
• Blood volume pressure
• Skin conductivity
• Respiration

Source http//www.univie.ac.atcga/courses/BE513/P
rojects/
18
Affective Signal Processing
Anger
Fear
Joy
Feature Extraction
Recognized Emotions
Emotion Name anger Emotion Intensity
0.6 Effective Threshold 0.2 Emotion Name
distress Emotion Intensity 0.7 Effective
Threshold 0.1
Psychological Signals
19
Cognitive Decision Making
Personality
Cognitive Behaviors
Releasers of Behaviors
Reactions
Emotions
Personality and Emotions influences the way
behaviors are exhibited and actions are performed.
20
Behavior ModelingProbabilistic Finite-state
Machines
Behavior Cell
Behavior Cell comprises of multiple (possibly
concurrent) behaviors.
21
Semantic Multimedia Organization
Learning content is organized in a hierarchical
structure by using a document object model
22
Some Assumptions

• Emotional status could be determined from
  psychological signals with acceptable accuracy
  higher than 75 have been reported.
• The solution is very cheap and easy to implement
  .
• The cognitive decision making process is
  influenced by knowledge, immediate emotional, the
  mood and personality factors of the learner.
• The cognitive decisions are non-deterministic,
  i.e. more than one learning sequence could be
  activated, but some decisions will be weakened
  for determinism based on preferences of the
  learner.

23
Thank You

• For more information please visit
• http//www.geocities.com/hekanayake/
• http//www.ucsc.cmb.ac.lk/People/hbe/

• Special Thanks,
• Supervisors Dr. D.D. Karunaratna Dr. K.P.
  Hewagamage
• NSF of Sri Lanka
• UCSC
• Individuals who supported

24
Additional
25
Biofeedback Sensing of Psychological Signals
A Simple LEGO Mindstorm-based GSR meter
Recording the GSR waveform (skin conductivity)
when a subject is observing an emotional
presentation