|
Title:
|
TOWARDS EVALUATING HUMAN-INSTRUCTABLE SOFTWARE AGENTS |
|
Author(s):
|
Robert D. Grant, David DeAngelis, Dan Luu, Dewayne E. Perry, Kathy Ryall |
|
ISBN:
|
978-972-8939-52-6 |
|
Editors:
|
Katherine Blashki |
|
Year:
|
2011 |
|
Edition:
|
Single |
|
Keywords:
|
User Study, Evaluation, Instructable Computing, Electronic Student, Machine Learning |
|
Type:
|
Full Paper |
|
First Page:
|
225 |
|
Last Page:
|
232 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
The Bootstrapped Learning (BL) project is an attempt to create software agents (e-students) that are instructable by human teachers through natural instruction methods [Oblinger, 2006]. In this paper, we present an introduction to BL and three years of case studies investigating the use of human subjects in evaluating e-students. In our studies we investigate human teachersÂ’ expectations of e-students and important differences between human and software learners, including the greater semantic understanding of humans, the eidetic memory of e-students, and the importance of various study parameters including timing issues and lesson complexity to human performance. |
|
|
|
|
|
|
