Transcript BigRedRoadShowPoster - Computer Science & Engineering

ILMDA:
Intelligent Learning Materials Delivery Agents
Login
Goal
The ILMDA project is aimed at building an intelligent agent
with machine learning capabilities to better deliver learning
materials to students.
The agent should be able to (1) self-configure its reasoning
process to decide which learning materials to deliver and
how to deliver them, (2) evaluate the learning materials in
terms of their appropriateness, and (3) assess the students’
performance.
New user panel
Approach Overview
Successful
Login
ILMDA Agent
Historical profile,
Real-time behavior
student
Parametric profile of
student and environment
Computer
lectures & GUI
Retrieval instructions
Profile updates
Statistics updates
ILMDA
Reasoning
Timely delivery
of examples &
exercise problems
Challenges
Pick tutorial panel
Tutorial panel
database
Examples
Exercise problems
Statistics
Corbett et al. (1999): “The arsenal of sophisticated
computational modules inherited from AI produce
learning gains of approximately .3 to 1.0 standard
deviation units compared with students learning the same
content in a classroom.”
Graesser et al. (2001): “Human tutors produce impressive
learning gains (between .4 and 2.3 standard deviation
units over classroom teachers), even though the vast
majority of tutors in a school’s system have modest
domain knowledge, have no training in pedagogical
techniques, and rarely use the sophisticated tutoring
strategies of ITSs.”
Graesser et al. (2001) criticize the current state of
tutoring systems: (1) If students merely keep guessing
until they find an action that gets positive feedback, they
can learn to do the right thing for the wrong reasons –
shallow learning; (2) The tutor does not ask students to
explain their actions; (3) The user interface of tutoring
systems lacks stepping back to see the “basic approach”;
and (4) When students learn quantitative skills, they are
usually not encouraged to see their work from a
qualitative, semantic perspective
Corbett, A., J. Anderson, A. Graesser, K. Koedinger, and K.
VanLehn (1999). Third Generation Computer Tutors: Learn from
or Ignore Human Tutors? in Proceedings of the 1999 Conference
of Computer-Human Interaction, 85-86.
Graesser, A. C., K. VanLehn, C. P. Rosé, P. W. Jordan, and D.
Harter (2001). Intelligent Tutoring Systems with Conversational
Dialogue, AI Magazine, 22(4):39-51.
Methodology
Our methodology (1) employs sound artificial intelligence (AI)
techniques such as case-based reasoning (CBR), reinforcement
learning, dynamic profiling, semantic search, rule-based reasoning,
simulated annealing, and so on, and (2) incorporates instructional
technology techniques such as adaptive quizzes, learning objects,
learner modeling, scaffolding, and so on, and (3) investigates how
agents can learn to better deliver learning materials to students.
Example panel
Problem panel
An intelligent agent that interacts with its environment (through its
GUI), makes decisions autonomously using CBR, and learns to
improve its performance.
It has a portable design with a Java-based GUI frontend and a
flexible design with a mySQL database backend.
Faculty
Leen-Kiat Soh
Suzette Person
Students
Todd Blank
L.D. Miller
Ashok Thirunavukkarasu
Contact Info
[email protected]
[email protected]
(402) 472-6738
(402) 472-1179
http://csce.unl.edu/agents/crush
Acknowledgements This work is supported in
part by Great Plains Software Technology Initiative,
National Center for Information Technology in
Education (NCITE), and Computer Science and
Engineering (CSE) at the University of Nebraska,
Lincoln.