Abstract
Activity recognition has become one of the emerging applications in the area of ubiquitous computing. This research aims at leveraging ubiquitous data stream mining and context reasoning for mobile activity recognition. The novel system allows dynamic adaptation and personalisation of the learning model to reflect the realistic activity changes emerged over time. Sensors fusion to attain a context aware activity recognition is also a key contribution in this project.
Original language | English |
---|---|
Title of host publication | Australasian Computing Doctoral Consortium (ACDC 2012), 30th January 2012 |
Subtitle of host publication | [part of the] Australasian Computer Science Week 2012 (ACSW 2012), 30 January - 3 February 2012, RMIT University, Melbourne, Australia |
Publisher | Australian Computer Society Inc |
Number of pages | 6 |
Publication status | Published - 2012 |
Keywords
- Activity recognition
- Context
- Adaptation
- Personalisation
- Stream learning
- Holistic approach