Metrics
4 File views/ downloads
56 Record Views
Thesis
THE APPLICATION OF STRUCTURAL CASE-BASED REASONING TO ACTIVITY RECOGNITION IN THE SMART HOME ENVIRONMENT
University of West Florida
Master of Science (MS), University of West Florida
2012
Abstract
Recent technological advances have made the smart home a potential alternative to health care facilities for providing eldercare. Designing a smart home requires the solution of two interrelated problems: designing a cost-effective sensor network that is capable of providing context information to an intelligent agent and designing the agent to interpret accurately only the relevant information from the incoming sensor data (Ye, Coyle, Dobson, & Nixon, 2009). In designing such an agent to recognize the activities being performed by the resident from an interpretation of the sensor data, case-based reasoning has a clear advantage over rule-based reasoning: whereas rules must be extracted, encoded, and maintained, experiences can simply be captured. We designed a smart home infrastructure with a sensor network to provide context data and a multi-agent intelligent middle layer, developed a Resource Description Framework (RDF) based ontology that incorporates constraint satisfaction for the temporal logic, and implemented a prototype of the case-based reasoning activity recognition agent. Initial results show that the agent recognizes the activities performed in our synthetic data set.
Details
- Title
- THE APPLICATION OF STRUCTURAL CASE-BASED REASONING TO ACTIVITY RECOGNITION IN THE SMART HOME ENVIRONMENT
- Resource Type
- Thesis
- Publisher
- University of West Florida
- Format
- text
- Identifiers
- 99380090749506600
- Academic Unit
- Computer Science
- Language
- English
- Awarding Institution
- University of West Florida; Master of Science (MS)
- Theses and Dissertations
- Master of Science (MS), University of West Florida