Temporal task footprinting
Authors:
Open Access
- 7 February 2010
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 281-284
Abstract
This paper introduces a new representation for describing routine tasks, called temporal task footprints. Routines are characterized by their temporal regularity or rhythm. Temporal pattern analysis (T-patterns) can be used to isolate frequent recurrent patterns in routine tasks that appear repeatedly in the same temporal configuration. Using tf-idf statistics, each task can then be defined in terms of its temporal task footprint, a ranked list of temporal patterns along with their typical frequencies. Experimental evaluations using data of 29 days observing and logging 10 subjects showed that temporal task footprints of application windows, email and document usage outperform decision tree and SVMs in recognizing the subjects' tasks.Keywords
This publication has 11 references indexed in Scilit:
- Discovering frequent work procedures from resource connections2009
- Frequent pattern mining: current status and future directionsData Mining and Knowledge Discovery, 2007
- A hybrid learning system for recognizing user tasks from desktop activities and email messages2006
- SWISH2006
- No task left behind?2005
- TaskTracer2005
- What a to-do2004
- Reconceptualizing Organizational Routines as a Source of Flexibility and ChangeAdministrative Science Quarterly, 2003
- Work rhythms2002
- A tutorial on hidden Markov models and selected applications in speech recognitionProceedings of the IEEE, 1989