Do users tolerate errors from their assistant?: experiments with an E-mail classifier

ABSTRACT

Smartlook, an e-mail classifier assistant, helps users filing their e-mails into folders. For a given message, it predicts the six most likely folders for that message and provides shortcut buttons that facilitate filing into one of the predicted folders. In this paper, we report results from user tests that show that although Smartlook does not achieve 100% prediction accuracy, a small percentage of errors does not hurt since users tolerate some errors from such an assistant.

REFERENCES

• 1: A. K. McCallum. Bow: A toolkit for statistical language modeling, text retrieval, classification and clustering, 1996. www.cs.cmu.edu/~mccallum/bow.
• 2: Richard Segal , Jeffrey O. Kephart, Incremental Learning in SwiftFile, Proceedings of the Seventeenth International Conference on Machine Learning, p.863-870, June 29-July 02, 2000
• 3: Ben Shneiderman, Looking for the bright side of user interface agents, interactions, v.2 n.1, p.13-15, Jan. 1995 [doi>10.1145/208143.208150]

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.5 PATTERN RECOGNITION
      I.5.2 Design Methodology
          Subjects: Classifier design and evaluation

Additional Classification:
  D. Software
  D.2 SOFTWARE ENGINEERING
      D.2.2 Design Tools and Techniques
          Subjects: User interfaces

  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.4 Systems and Software

          Nouns: World Wide Web (WWW)
  H.4 INFORMATION SYSTEMS APPLICATIONS
      H.4.3 Communications Applications
          Subjects: Electronic mail

General Terms:
Design, Human Factors, Performance, Theory

Keywords:
E-mail classification, personal assistant, usability