1st International IEEE Conference on Pervasive Services

Research Article

Path Prediction through Data Mining

  • @INPROCEEDINGS{10.1109/PERSER.2007.4283902,
        author={Theodoros Anagnostopoulos and Christos  B. Anagnostopoulos and Stathes  Hadjiefthymiades and Alexandros Kalousis and Miltos  Kyriakakos},
        title={Path Prediction through Data Mining},
        proceedings={1st International IEEE Conference on Pervasive Services},
        publisher={IEEE},
        proceedings_a={ICPS},
        year={2007},
        month={8},
        keywords={data mining  location prediction  machine learning},
        doi={10.1109/PERSER.2007.4283902}
    }
    
  • Theodoros Anagnostopoulos
    Christos B. Anagnostopoulos
    Stathes Hadjiefthymiades
    Alexandros Kalousis
    Miltos Kyriakakos
    Year: 2007
    Path Prediction through Data Mining
    ICPS
    IEEE
    DOI: 10.1109/PERSER.2007.4283902
Theodoros Anagnostopoulos1,*, Christos B. Anagnostopoulos1,*, Stathes Hadjiefthymiades1,*, Alexandros Kalousis2,*, Miltos Kyriakakos1,*
  • 1: Pervasive Computing Research Group, Communication Networks Laboratory, Department of Informatics and Telecommunications, University of Athens, Panepistimiopolis, Ilissia, Athens 15784, Greece, tel: +302107275127
  • 2: Artificial Intelligence Laboratory, Department of Computer Science, University of Geneva, Uni-Dufour, Geneva 1211, Switzerland, tel: +41223797630
*Contact email: thanag@di.uoa.gr, bleu@di.uoa.gr, shadj@di.uoa.gr, Alexandros.Kalousis@cui.unige.ch, miltos@di.uoa.gr

Abstract

Context-awareness is viewed as one of the most important aspects in the emerging ubiquitous computing paradigm. However, mobile applications are required to operate in pervasive computing environments of dynamic nature. Such applications predict the appropriate context in their environment in order to act efficiently. A context model, which deals with the location prediction of moving users, is proposed. Such model is used for trajectory classification through machine learning techniques. Hence, spatial and spatiotemporal context prediction is regarded as context classification based on supervised learning. Finally, two classification schemes are presented, evaluated and compared with other ML schemes in order to support location prediction and decision making.