Transcript ITAC2003 - Department of Computer Science and Engineering

Developing Intelligent Agents and
Multiagent Systems for
Educational Applications
Leen-Kiat Soh
Department of Computer Science and Engineering
University of Nebraska
[email protected]
What is an Agent?
• An agent is an entity that takes sensory input
from its environment, makes autonomous
decisions, and carries out actions that affect the
environment
Agent
sensory
input
think!
Environment
Agents and MAS | I-MINDS | ILMDA
output
actions
What is an Intelligent Agent?
• An intelligent agent is one that is capable of
flexible autonomous actions in order to meet its
design objectives, where flexibility means:
– Reactivity, Pro-activeness and Social ability
• Machine Learning in AI says
The acquisition of new knowledge and motor and cognitive skills
and the incorporation of the acquired knowledge and skills in
future system activities, provided that this acquisition and
incorporation is conducted by the system itself and leads to an
improvement in its performance.
• Not all agents are intelligent!
Agents and MAS | I-MINDS | ILMDA
What is a Multiagent System?
• A multiagent system is a system where multiple
agents perform a task better when working
together
– Interaction (communication)
– Coordination
– Collaboration
• Example: A group of basketball players who do
not observe or communicate with each other is
not a team—simply a group of individual agents.
Agents and MAS | I-MINDS | ILMDA
Education Systems
• Not all computer-aided learning and teaching
systems are agent-based, not all are intelligent
• Systems related to agents and multiagent
systems focus on three areas:
– Intelligent User Interface
– Tutors
– Multiagent Systems
Agents and MAS | I-MINDS | ILMDA
Intelligent Multiagent Infrastructure for
Distributed Systems in Education (I-MINDS)
Agents and MAS | I-MINDS | ILMDA
I-MINDS: Goals
• To build a multiagent infrastructure for
distributed systems, in an education application
– To employ multiagent intelligence to facilitate
teaching and learning processes
– To enhance peer (or collaborative) learning among
students
– To loosen spatial and temporal constraints of
conventional lecture delivery (for distance learning)
Agents and MAS | I-MINDS | ILMDA
Help Teachers Teach
• I-MINDS helps teachers teach
–
–
–
–
–
–
–
–
–
Organize lectures (distance learning, e-archival)
Manage students
Keep track of classroom activities
Profile students dynamically
Rank real-time questions
Help deliver customized questions/quizzes/homework
Learn about the students
Learn about the lectures
Learn about the ranking of questions (keywords)
Agents and MAS | I-MINDS | ILMDA
Help Students Learn
• I-MINDS helps students learn
–
–
–
–
–
–
–
–
Organize lectures (distance learning, e-archival)
Keep track of classroom activities
Profile students dynamically
Form “buddy group” (peer learning and real-time
collaboration via forum and whiteboard)
Encourage students to ask questions and to be more
proactive
Learn about forming buddy group dynamically
Learn about good questions
Learn about good answers
Agents and MAS | I-MINDS | ILMDA
Capabilities (Intelligent Agents and MAS)
Teacher Agent
•
•
•
•
evaluate each (textual) question (text) based on its timestamp,
content, images, quality (keyword-based), and the profile of the
questioner
rank audio questions based on student profile
profile each student based on the number of questions asked,
number of questions answered by the teacher, average length
and quality of questions
communicate, transmit lectures, archive, collect statistics,
monitor the system
Student Agent
•
•
•
profile each student in the buddy group based on their
response
adaptively refine the buddy group based on the buddies’ profile
communicate, transmit questions and responses, archive,
collect statistics, monitor the system
Agents and MAS | I-MINDS | ILMDA
Screenshots
Agents and MAS | I-MINDS | ILMDA
Screenshots
Agents and MAS | I-MINDS | ILMDA
Intelligent Learning Materials Delivery Agent
(ILMDA)
Agents and MAS | I-MINDS | ILMDA
ILMDA: Goals
• Build an intelligent agent with machine learning
capabilities to deliver better learning materials to
students
– Incorporates instructional technology techniques such
as adaptive quiz, learning objects, learner modeling,
and so on
– Investigates how agents can learn to deliver better
learning materials to students
– Employs sound artificial intelligence (AI) techniques
• case-based reasoning, reinforcement learning, dynamic
profiling, semantic search, rule-based reasoning, simulated
annealing
Agents and MAS | I-MINDS | ILMDA
Design
• ILMDA delivers learning materials based
on
– The usage history of the learning materials
• Each Learning Material consists of a tutorial, a set
of related examples, and a set of exercise
problems
– The student static background profile
• E.g., GPA, majors, interests
– The student dynamic activity profile
• Based on their interactions with the agent
Agents and MAS | I-MINDS | ILMDA
Assumptions
• Assumption 1: A student’s behavior in viewing
an online tutorial, and how he or she interacts
with the tutorial, the examples, and the
exercises
– is a good indicator of how well the student is
understanding the topic in question, and
– this behavior is observable and quantifiable
• Assumption 2: Different students exhibit
different behaviors for different topics
consistently enough to be recognized as patterns
Agents and MAS | I-MINDS | ILMDA
Helps Students Learn
Historical profile,
Real-time behavior
Parametric profile of
student and environment
Retrieval instructions
Profile updates
Statistics updates
ILMDA
Reasoning
student
lectures
Computer
& GUI
Timely delivery
of examples &
exercise problems
Examples
database
Exercise problems
Statistics
ILMDA Agent
Agents and MAS | I-MINDS | ILMDA
Capabilities
Front-End GUI
• To register students, capture student dynamic profile,
deliver learning materials,
• To enter learning materials
• To enter domain expertise – heuristics, weights
Intelligent Agent
• CBR, multi-layered learning modules, database
retrieval, self-evaluation
Backend Database
• mySQL database, multiple databases for content,
expertise, profiles
Agents and MAS | I-MINDS | ILMDA
Screenshots
Agents and MAS | I-MINDS | ILMDA
Screenshots
Agents and MAS | I-MINDS | ILMDA
To Probe Further
@
http://www.cse.unl.edu/agents