high threshold values, which results in high FN(%)
values. But ATiS achieves low FP(%) and FN(%) values
simultaneously. However, the values differ for every
individual user and on a daily basis as shown in figure 6 (b). Overall, we achieved average FP(%) and
FN(%) values of 1. 10 percent and 0.19 percent, which
is very close to the ideal case of zero FP(%) and FN(%)
We use a digital power monitoring device from
Monsoon Solutions4 to measure the energy consumptions for event sensing on Android smartphones (Google Nexus One). Extensive trials are
done to avoid sensitive fluctuations in power consumption. Table 3 shows the general energy consumptions per second for location sensing by our system and other available techniques. However, the
total amount of energy consumption varies differently depending on the application scenarios.
The Platys project builds on a semantic concept of
place to facilitate developing context-aware mobile
applications that can enhance their users’ experi-
ence. A place in Platys goes beyond location to
include associated time spans, activities, people,
roles, and objects. Our resulting context model is
supported by an ontology in OWL.
Place recognition is performed using a semisupervised EM algorithm as well as standard machine-learning classifiers. Our approach allows to capture
nuances in how a user perceives places and is able to
recognize ( 1) user’s place and activity at different levels of granularity; ( 2) disjoint spatial regions as a single place; and ( 3) the same spatial region as more
than one place (for the same user and for different
users). Performance for recognizing place at a general level (at home versus at work versus elsewhere)
using machine-learning classifiers is higher than that
reported in existing works. Performance for recognizing place using a semisupervised EM algorithm
was generally better than two stay-point approaches
used for comparison. Location plays an important
role in place recognition. We have addressed the
problem of energy-efficient location sensing.
A place is naturally associated with a social context. We have proposed an approach to recognize
social circles by exploiting place information. Our
approach performs best when places are defined
using both spatial and social attributes.
To provide users with privacy to protect the personal information their mobile devices are collecting,
we define privacy and information sharing policies.
The policies are expressed in the SemanticWeb languages OWL and RDF. Our policies ensure context-dependent release and obfuscation of information in
accordance to the user preferences.
This material is based upon work supported by the
National Science Foundation under Grants 0910838,
0910868, 0910846, and 1016216.
Baader, F.; Calvanese, D.; McGuinness, D. L.; Nardi, D.; and
Patel-Schneider, P. F., eds. 2003. The Description Logic Hand-
book: Theory, Implementation, and Applications. New York:
Cambridge University Press.
Baldauf, M.; Dustdar, S.; and Rosenberg, F. 2007. A Survey
on Context-Aware Systems. International Journal of Ad Hoc
and Ubiquitous Computing
2( 4): 263–277. dx.doi.org/10.
Bao, L., and Intille, S. S. 2004. Activity Recognition from
User-Annotated Acceleration Data. In Pervasive Computing,
Lecture Notes in Computer Science volume 3001, 1–17.
Berlin: Springer. dx.doi.org/10.1007/978-3-540-24646-6_1
Bechhofer, S.; van Harmelen, F.; Hendler, J.; Horrocks, I.;
Mcguinness, D. L.; Patel-Schneider, P. F.; and Stein, L. A.
2007. Owl Web Ontology Language Reference. Technical
Report, W3C. Boston, MA: The World Wide Web Consortium.
Carroll, J. J.; Dickinson, I.; Dollin, C.; Reynolds, D.;
Seaborne, A.; and Wilkinson, K. 2004. Jena: Implementing
the Semantic Web Recommendations. In Proceedings of the
13th International World Wide Web Conference,
York: Association for Computing Machinery.
Chen, G., and Kotz, D. 2000. A Survey of Context-Aware
Mobile Computing Research. Technical Report, Dartmouth
College Computer Science Department, Hanover, NH.
Chen, H.; Finin, T.; and Joshi, A. 2005. The SOUPA Ontology
for Pervasive Computing, 233–258. Berlin: Birkhauser Publishing Ltd. / Springer.
Dey, A. K.; Abowd, G. D.; and Salber, D. 2001. A Conceptual Framework and a Toolkit for Supporting the Rapid Prototyping of Context-Aware Applications. Human-Computer
16( 2): 97–166. dx.doi.org/10.1207/S15327051
Eagle, N., and Pentland, A. 2006. Reality Mining: Sensing
Complex Social Systems. Personal and Ubiquitous Computing
Table 3. Energy Consumption Per Second Between Our System
and Other Techniques for Continuous Location Sensing.
Item Energy Consumed (m Wh)
Our System 0.0173
Wi-Fi Scan 0.1185