Explanation of adaptive systems
Abstract/Contents
- Abstract
- Software systems are becoming increasingly complex. Common everyday applications like word processors and email clients often contain sophisticated algorithms for classifying inputs, suggesting improvements, adjusting to changing conditions, and predicting user intent. Even traditionally simple systems are increasingly adding adaptive components, in which the system learns and adapts its behavior over time. Despite this increasing sophistication, however, these systems typically provide little transparency into the computation and reasoning being performed. These complex systems are putting an increasing burden on end users to trust their reasoning and computation, but seldom provide the tools and interfaces necessary for building this trust. We investigate how context-sensitive explanatory information provided to users by adaptive systems can help to establish and build confidence in computational results and conclusions. We then describe the Integrated Cognitive Explanation Environment (ICEE), an explanation framework we designed and implemented that addresses the issues involved with interacting with adaptive systems. This framework provides a uniform approach for representing and explaining both provenance and inference information related to logical deduction, task processing, and machine learning. For this framework, there is particular emphasis on the information that systems need to provide in order to be considered "explainable." We also show how the framework can be used for a variety of computational tasks, using a single unified representation. Finally, we explore the implications of adaptive systems that support a complex context-sensitive explanation system. We focus primarily on how the use of explanations can be used to enhance machine learning in adaptive systems.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2011 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Glass, Alyssa |
|---|---|
| Associated with | Stanford University, Computer Science Department |
| Primary advisor | Genesereth, Michael R, 1948- |
| Thesis advisor | Genesereth, Michael R, 1948- |
| Thesis advisor | McGuinness, Deborah L |
| Thesis advisor | Nass, Clifford Ivar |
| Advisor | McGuinness, Deborah L |
| Advisor | Nass, Clifford Ivar |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Alyssa Glass. |
|---|---|
| Note | Submitted to the Department of Computer Science. |
| Thesis | Ph.D. Stanford University 2011 |
| Location | electronic resource |
Access conditions
- Copyright
- © 2011 by Alyssa Glass
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...