Discovery of activity composites using topic models: An analysis of unsupervised methods. Seiter, J., Amft, O., Rossi, M., & Tröster, G. Pervasive and Mobile Computing, 15:215–-227, December, 2014.
doi  abstract   bibtex   
In this work we investigate unsupervised activity discovery approaches using three topic model (TM) approaches, based on Latent Dirichlet Allocation (LDA), n-gram TM (NTM), and correlated TM (CTM). While LDA structures activity primitives, NTM adds primitive sequence information, and CTM exploits co-occurring topics. We use an activity composite/primitive abstraction and analyze three public datasets with different properties that affect the discovery, including primitive rate, activity composite specificity, primitive sequence similarity, and composite-instance ratio. We compare the activity composite discovery performance among the TM approaches and against a baseline using k-means clustering. We provide guidelines for method and optimal TM parameter selection, depending on data properties and activity primitive noise. Results indicate that TMs can outperform kk-means clustering up to 17%, when composite specificity is low. LDA-based TMs showed higher robustness against noise compared to other TMs and k-means.
@Article{Seiter2014-J_PervasiveMobComput,
  Title                    = {Discovery of activity composites using topic models: An analysis of unsupervised methods},
  Author                   = {Julia Seiter and Oliver Amft and Mirco Rossi and Gerhard Tr\"{o}ster},
  Journal                  = {Pervasive and Mobile Computing},
  Year                     = {2014},

  Month                    = {December},
  Pages                    = {215–-227},
  Volume                   = {15},

  Abstract                 = {In this work we investigate unsupervised activity discovery approaches using three topic model (TM) approaches, based on Latent Dirichlet Allocation (LDA), n-gram TM (NTM), and correlated TM (CTM). While LDA structures activity primitives, NTM adds primitive sequence information, and CTM exploits co-occurring topics. We use an activity composite/primitive abstraction and analyze three public datasets with different properties that affect the discovery, including primitive rate, activity composite specificity, primitive sequence similarity, and composite-instance ratio. We compare the activity composite discovery performance among the TM approaches and against a baseline using k-means clustering. We provide guidelines for method and optimal TM parameter selection, depending on data properties and activity primitive noise. Results indicate that TMs can outperform kk-means clustering up to 17\%, when composite specificity is low. LDA-based TMs showed higher robustness against noise compared to other TMs and k-means.},
  Doi                      = {10.1016/j.pmcj.2014.05.007},
  File                     = {Seiter2014-J_PervasiveMobComput.pdf:Seiter2014-J_PervasiveMobComput.pdf:PDF},
  Keywords                 = {Activity routines; Daily routines; Topic modeling; Hierarchical activity recognition; Activity discovery},
  Owner                    = {oamft},
  Timestamp                = {2013/09/16}
}

Downloads: 0