WPI Worcester Polytechnic Institute

Computer Science Department
------------------------------------------

AIRG Topics - Fall 2005


Our group meets on Thursdays at 11:00 a.m., FL 246, Beckett Conf. Room.


Dates and topics for this semester are as follows:

Sept 8
AIRG Organizational Meeting (Coordinator: DCB)

Sept 15
Mike Macasek
Intelligent Tutoring System demonstration

Sept 22
Dave Brown
The Relationship between Function and Affordance

Sept 29
Mingyu Feng
Longitudinal Data Analysis on Assistment Data

Oct 6
Iain Boyle
CAFixD: A case-based reasoning fixture design method

Oct 13
Marc DiNino, Jim Kazmierczak, Mary Plunkett
WITS Tutoring system MQP
(in cooperation with SEU and NASA GSFC)

Oct 20
{no mtg}

Oct 27
Aparna Varde
LearnMet: Learning Domain-Specific Distance Metrics for Plots of Scientific Functions
PhD Progress report
    ABSTRACT: Scientific experimental results are often depicted as plots of functions to aid their visual analysis and comparison. In computationally comparing these plots using techniques such as similarity search and clustering, the notion of similarity is typically distance. However, it is seldom known which distance metric(s) best preserve(s) semantics in the given domain. It is thus desirable to learn such metrics for domain-specific comparison of plots. This is the goal of our proposed technique, LearnMet. The input to LearnMet is a training set with correct clusters of plots. These are iteratively compared with clusters over the same plots obtained using an arbitrary but fixed clustering algorithm. Using a guessed initial metric for clustering, adjustments are made to the metric in each epoch based on the error between the obtained and correct clusters until the error is minimal or below a given threshold. The metric giving the lowest error is the learned metric. The proposed LearnMet technique and its refinements are discussed in this paper. The primary application of LearnMet is clustering plots in the Heat Treating domain. Hence it is rigorously evaluated using Heat Treating data. Given distinct test sets for evaluation, clusters of plots predicted using the learned metrics are compared with given actual clusters over the same plots. The extent to which the predicted and actual clusters match each other denotes the quality of the learned metrics.

Nov 3
Terrence Turner
Intelligent Tutoring
(MS Thesis presentation)

Nov 10
Goss Nuzzo-Jones
Common Tutor Object Platform - An Intelligent Tutoring System Software Development Strategy
(MS Thesis presentation)

Nov 17
John Hayward
Mining Oncological Data
(MS Thesis proposal presentation)

Nov 24
Thanksgiving Break

Dec 1
Jay Walonoski
Gaming Detection and Prevention in Intelligent Tutoring Systems
(MS Thesis presentation)
    ABSTRACT: A major issue in Intelligent Tutoring Systems is off-task student behavior, especially performance-based gaming, where students systematically exploit tutor behavior in order to advance through a curriculum quickly and easily, with as little active thought directed at the educational content as possible. This talk will focus on two main areas of research related to gaming behavior, using the Assistments system as the experimental test bed. The first area is the detection of gaming behavior within the Assistments system, using machine learning techniques and methods pioneered in prior studies in related systems. The second area is our initial attempt to develop a non-intervening mechanism for the determent and prevention of off-task gaming behavior.

Dec 8
Kevin Menard
A study of voluntary feedback
(in cooperation with Microsoft)

Dec 15
No meeting


[Feedback] [Search Our Web] [Help & Index]

[Return to the WPI Homepage] [Return to the CS Homepage]

AIRG Coordinator / Wed Oct 26 18:20:22 EDT 2005