Transcript mitrovic
Intelligent Tutors for All:
the Constraint-based
Approach
Tanja Mitrovic
Intelligent Computer Tutoring Group
University of Canterbury
Intelligent Tutoring Systems
Goal: one-to-one teaching without the expense of
human tutoring
Simulate a human teacher
Problem-solving environments (learning by doing)
Based on Artificial Intelligence
Student modeling
Student modeling
Data about user
System
Student
Model
Adaptation Effect
Architecture of ITSs
Domain
knowledge
Domain
module
Student
modeler
Pedagogical
module
Pedagogical
expertise
Student
Models
Interface
Communication
knowledge
Student
Learning from performance errors
Ohlsson, 1992
Declarative/procedural knowledge
Constraints as a knowledge-representation formalism
Constraints do not assert anything
Constraints encode correctness for a domain
“If the relevance condition R is true,
then the satisfaction condition S ought to be true,
otherwise something is wrong.”
Constraints support judgment, not inference
Learning from performance errors
Learning phases:
Error detection
Error correction
How can we catch ourselves making errors?
If the knowledge is there, then why the error?
If not, then how is the error detected?
CBM: domain and student modeling
Constraint-based Modeling
The space of incorrect knowledge is vast
Therefore: abstractions are needed
Represent only basic domain principles
Group the states into equivalence classes
according to their pedagogical importance
Constraint-Based Modeling
Domain knowledge represented by a set of
constraints
A constraint is a pattern of form <Cr, Cs>
If a solution matches the Cr then it must also
match the Cs, else something is wrong
“Innocent until proven guilty” approach
Example constraints
If you are driving in New Zealand,
you better be on the left side of the road.
If the current problem is a/b + c/d,
and the student’s solution is (a+c)/n,
then it had better be the case that n=b=d.
Advantages of CBM
Very efficient computationally
No need for a problem solver
No need for a bug library
Insensitive to the radical strategy variability
phenomenon
Neutral with respect to pedagogy
Implications for ITS Design: CBM
Represent the domain in terms of constraints
Model the student in terms of constraints
Pedagogy:
Augment student’s constraint base
When should the ITS take an initiative?
What to instruction to deliver?
Models of meta-cognitive skills
Student’s meta-cognitive skills
CBM: Model the Student
A violated constraint implies incomplete or
incorrect knowledge
Short-term student model:
the set of violated constraints
the set of satisfied constraints
No one-to-one mapping between problems and
constraints
Long-term student model:
Constraint histories (overlay/probabilistic)
CBM: Pedagogy
Constraint-based tutors function by augmenting
the student’s own knowledge base
Choose practice problems that exercise
constraints
Interrupt when a constraint is violated
Attach feedback messages to the constraints
Tell the student which constraint he/she just
violated and how
History of ICTG
SQL-Tutor
Solaris (1997), Windows (1998), Web (1999)
CAPIT (2000)
KERMIT (2000)
WETAS (2002)
LBITS (2002)
NORMIT (2002)
ERM-Tutor (2003)
COLLECT-UML (2005)
ASPIRE, VIPER
J-LATTE
Thermo-Tutor
CAPIT
LBITS – elementary vocabulary
Copy
Paste
Pen
Get the pen, each
time you want to
update the group
diagram and
Leave it as soon
as you are done
Group Diagram
Chat Area
Individual Diagram
Feedback Area
Current work
ASPIRE, VIPER
Supporting meta-cognitive skills (selfexplanation, self-assessment …)
Affective modeling and pedagogical agents
Supporting multiple teaching strategies
New ITSs