Mobile and Ubiquitous Systems: Computing, Networking, and Services. 7th International ICST Conference, MobiQuitous 2010, Sydeny, Australia, December 6-9, 2010, Revised Selected Papers

Research Article

Mining Emerging Sequential Patterns for Activity Recognition in Body Sensor Networks

Download60 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-29154-8_9,
        author={Tao Gu and Liang Wang and Hanhua Chen and Guimei Liu and Xianping Tao and Jian Lu},
        title={Mining Emerging Sequential Patterns for Activity Recognition in Body Sensor Networks},
        proceedings={Mobile and Ubiquitous Systems: Computing, Networking, and Services. 7th International ICST Conference, MobiQuitous 2010, Sydeny, Australia, December 6-9, 2010, Revised Selected Papers},
        proceedings_a={MOBIQUITOUS},
        year={2012},
        month={10},
        keywords={Body sensor networks activity recognition data mining},
        doi={10.1007/978-3-642-29154-8_9}
    }
    
  • Tao Gu
    Liang Wang
    Hanhua Chen
    Guimei Liu
    Xianping Tao
    Jian Lu
    Year: 2012
    Mining Emerging Sequential Patterns for Activity Recognition in Body Sensor Networks
    MOBIQUITOUS
    Springer
    DOI: 10.1007/978-3-642-29154-8_9
Tao Gu1,*, Liang Wang,*, Hanhua Chen,*, Guimei Liu2,*, Xianping Tao3,*, Jian Lu3,*
  • 1: University of Southern Denmark
  • 2: National University of Singapore
  • 3: Nanjing University
*Contact email: gu@imada.sdu.dk, wang@imada.sdu.dk, hhchen@imada.sdu.dk, liugm@comp.nus.edu.sg, txp@nju.edu.cn, lj@nju.edu.cn

Abstract

Body Sensor Networks offer many applications in healthcare, well-being and entertainment. One of the emerging applications is recognizing activities of daily living. In this paper, we introduce a novel knowledge pattern named Emerging Sequential Pattern (ESP)—a sequential pattern that discovers significant class differences—to recognize both simple (i.e., sequential) and complex (i.e., interleaved and concurrent) activities. Based on ESPs, we build our complex activity models directly upon the sequential model to recognize both activity types. We conduct comprehensive empirical studies to evaluate and compare our solution with the state-of-the-art solutions. The results demonstrate that our approach achieves an overall accuracy of 91.89%, outperforming the existing solutions.