|Semantic Web; RDF; Conversion; Smart home; Indoor localization; RFID
|Full text PDF:
The Semantic Web, which is a machine understandable Web likely to be the future of the Web, is being developed and capturing much attention in recent years. Many things need to be done to ensure the success of moving from the Syntactic Web (the current Web) to the Semantic Web. The first problem identified is there being considerable amounts of data being stored as XML documents for the current Web. However, as XML was invented for structuring data rather than describing the meaning of data, XML documents cannot be used as they are in the Semantic Web environment. One way to make the XML data usable to the Semantic Web is converting them to RDF formatted data. However, it is labor-intensive to perform the conversion manually. The second problem relates to building a Smart Home system that will operate in the Semantic Web environment and assist the occupier. For example, people sometimes forget where they have left certain things, such as glasses, wallets, keys, etc. and it is often troublesome to find them. This research aims to address the two identified problems. For the first one, a universal XML to RDF conversion on a large scale algorithmic solution is developed to automatically transform XML documents to RDF documents. Outputs of this solution include a transformation procedure and an actual implementation of the procedure in the form of a Java tool called X2R (XML to RDF). For the second identified problem, this study proposes an RFID solution to localize the easily-lost objects based on three nearest reference points and a recommendation as to how to build the indoor localization system, named HLSM (Home Localization System for Misplaced objects), so that it can be easily integrated into the entire Smart Home system that adheres to the Semantic Web context. Promising results were obtained for both solutions to the two aforementioned problems. The universal XML to RDF conversion solution can efficiently transform XML documents to RDF with 100% accuracy. One could confidently apply the X2R tool to convert XML documents on a large scale. In regard to the other solution, the HLSM system prototype provided a localization accuracy rate of 87.5%, in the form of user-friendly statements, e.g. “The glasses are at the kitchen table”, which are not given by existing solutions. These encourage further enhancement of the localization technique and the implementation of the complete HLSM system.