Interacting meaningfully with machine learning systems : three experiments Public Deposited

http://ir.library.oregonstate.edu/concern/technical_reports/j3860812v

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Although machine learning is becoming commonly used in today's software, there has been little research into how end users might interact with machine learning systems, beyond communicating simple "right/wrong" judgments. If the users themselves could somehow work hand-in-hand with machine learning systems, the accuracy of learning systems could be improved and the users' understanding and trust of the system could improve as well. We conducted three experiments to begin to understand the potential for rich interactions between users and machine learning systems. The first experiment was a think-aloud study, aiming to see how willing users were to interact with and about machine learning reasoning, and to help us understand what kinds of feedback users might give to machine learning systems. Specifically, users were shown explanations of machine learning predictions and asked to provide feedback to improve the predictions. The results were that users' feedback was rich, complex, and widely varied, ranging from suggestions for reweighting of features to proposals for new features, feature combinations, relational features, and wholesale changes to the learning algorithm. We then investigated the viability of introducing such feedback into machine learning systems: specifically, how to incorporate some of these types of user feedback into machine learning systems, and impact on the accuracy of the system. Taken together, the results of our experiments show that supporting rich interactions between users and machine learning systems is feasible for both user and machine. This shows the potential of rich human-computer collaboration via on-the-spot interactions as a promising direction for machine learning systems to work more intelligently, hand-in-hand with the user.
Resource Type
Date Available
Date Issued
Series
Subject
Rights Statement
Publisher
Peer Reviewed
Language
Replaces
Additional Information
  • description.provenance : Submitted by Laura Wilson (laura.wilson@oregonstate.edu) on 2012-12-26T22:24:01Z No. of bitstreams: 1 2007-46.pdf: 387207 bytes, checksum: e504f44cb0ce0a98e9a39a636dd61d9d (MD5)
  • description.provenance : Approved for entry into archive by Laura Wilson(laura.wilson@oregonstate.edu) on 2012-12-26T22:25:18Z (GMT) No. of bitstreams: 1 2007-46.pdf: 387207 bytes, checksum: e504f44cb0ce0a98e9a39a636dd61d9d (MD5)
  • description.provenance : Made available in DSpace on 2012-12-26T22:25:18Z (GMT). No. of bitstreams: 1 2007-46.pdf: 387207 bytes, checksum: e504f44cb0ce0a98e9a39a636dd61d9d (MD5) Previous issue date: 2007

Relationships

In Administrative Set:
Last modified: 07/18/2017

Downloadable Content

Download PDF
Citations:

EndNote | Zotero | Mendeley

Items