APE: learning user's habits to automate repetitive tasks

ABSTRACT

The APE (Adaptive Programming Environment) project focuses on applying Machine Learning techniques to embed a software assistant into the VisualWorks Smalltalk interactive programming environment. The assistant is able to learn user's habits and to automatically suggest to perform repetitive tasks on his behalf. This paper describes our assistant and focuses more particularly on the learning issue. It explains why state-of-the-art Machine Learning algorithms fail to provide an efficient solution for learning user's habits, and shows, through experiments on real data that a new algorithm we have designed for this learning task, achieves better results than related algorithms.

REFERENCES

• 1: R. Armstrong, D. Freitag, T. Joachims, and T. Mitchell. Webwatcher: A learning apprentice for the world wide web. In AAAI Spring Symposium on Information Gathering, 1995.
• 2: A. Caglayan, M. Snorrason, J. Jacoby, J. Mazzu, and R. J. and. K. Kumar. Learn sesame: a learning agent engine. Applied Artificial Intelligence, 11:393-412, 1997.
• 3: Allen Cypher , Daniel C. Halbert , David Kurlander , Henry Lieberman , David Maulsby , Brad A. Myers , Alan Turransky, Watch what I do: programming by demonstration, MIT Press, Cambridge, MA, 1993
• 4: Richard M. Karp , Raymond E. Miller , Arnold L. Rosenberg, Rapid identification of repeated patterns in strings, trees and arrays, Proceedings of the fourth annual ACM symposium on Theory of computing, p.125-136, May 01-03, 1972, Denver, Colorado, United States [doi>10.1145/800152.804905]
• 5: Pattie Maes, Agents that reduce work and information overload, Communications of the ACM, v.37 n.7, p.30-40, July 1994 [doi>10.1145/176789.176792]
• 6: Tom M. Mitchell , Rich Caruana , Dayne Freitag , John McDermott , David Zabowski, Experience with a learning personal assistant, Communications of the ACM, v.37 n.7, p.80-91, July 1994 [doi>10.1145/176789.176798]
• 7: Tom Michael Mitchell, Version spaces: an approach to concept learning., 1979
• 8: H. Motoda. Machine Learning Techniques to Make Computers Easier to Use. In Proceedings of IJCAI'97. Morgan Kaufmann Publishers, August 23-29, 1997.
• 9: T. Niblett. Constructing decision trees in noisy domains. In I. Bratko and N. Lavraec, editors, Progress in Machine Learning, pages 67-78, Wilmslow, 1987. Sigma Press.
• 10: J. Ross Quinlan, C4.5: programs for machine learning, Morgan Kaufmann Publishers Inc., San Francisco, CA, 1993
• 11: J.-D. Ruvini and C. Fagot. IBHYS: a new approach to learn users habits. In Proceedings of ICTAI'98, pages 200-207. IEEE Computer Society Press, 1999

INDEX TERMS

Primary Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE

Additional Classification:
  I. Computing Methodologies
  I.2 ARTIFICIAL INTELLIGENCE

General Terms:
Experimentation, Human Factors, Management

Keywords:
interactive programming envionments, learning interface agents, machine learning, programming by demonstration