=Paper=
{{Paper
|id=Vol-1670/paper-69
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1670/paper-69.pdf
|volume=Vol-1670
}}
==None==
<pdf width="1500px">https://ceur-ws.org/Vol-1670/paper-69.pdf</pdf>
<pre>
                  Mining Subgroups with
              Exceptional Transition Behavior
                                   (Abstract)

           Florian Lemmerich1,5 , Martin Becker2 , Philipp Singer1,5 ,
          Denis Helic3 , Andreas Hotho2,4 , and Markus Strohmaier1,5
                            1
                             GESIS, Cologne, Germany
      {florian.lemmerich,philipp.singer,markus.strohmaier}@gesis.org
                       2
                         University of Würzburg, Germany
                {becker,hotho}@informatik.uni-wuerzburg.de
                 3
                    Graz University of Technology, Graz, Austria
                                dhelic@tugraz.at
                   4
                     L3S Research Center, Hannover, Germany
               5
                  University of Koblenz-Landau, Mainz, Germany



      Abstract. In this talk, we present a recently developed method for de-
      tecting interpretable subgroups with exceptional transition behavior in
      sequential data [1]. Potential applications of this technique include, e.g.,
      studying human mobility or analyzing the behavior of internet users. To
      approach this task, we extend exceptional model mining. Exceptional
      model mining provides a framework for mining interpretable data sub-
      sets with unusual interactions between a set of target attributes consid-
      ering a user-chosen model class. However, previously investigated model
      classes cannot capture transition behavior. Thus, we introduce first-order
      Markov chains as a novel model class for exceptional model mining and
      present a new interestingness measure that quantifies the exceptionality
      of transition subgroups. The measure compares the distance between the
      Markov transition matrix of a subgroup and the respective matrix of the
      entire data with the distance of random dataset samples. In addition, our
      method can be adapted to find subgroups that match or contradict given
      transition hypotheses. We demonstrate that our method is consistently
      able to recover subgroups with exceptional transition models from syn-
      thetic data and illustrate its potential in two application examples. Our
      work is relevant for researchers and practitioners interested in detecting
      exceptional transition behavior in sequential data.


References
1. Lemmerich, F., Becker, M., Singer, P., Helic, D., Hotho, A., Strohmaier, M.: Min-
   ing subgroups with exceptional transition behavior. In: Proceedings of the 21th
   ACM SIGKDD International Conference on Knowledge Discovery and Data Min-
   ing (2016), http://www.kdd.org/kdd2016/papers/files/Paper 185.pdf

</pre>