No abstract

To manipulate semantic web and integrate different data sources efficiently, automatic schema matching plays a key role. A generic schema matching method generally includes two phases: the linguistic similarity matching phase and the structural similarity matching phase. Since linguistic matching is an essential step for effective schema matching, developing a high accurate linguistic similarity matching scheme is required. In this paper, a schema matching approach called Similarity Yield Matcher (SYM) is proposed. In SYM, a lexical decision tree is presented to determine the linguistic similarity matching of the first phase. A structural matching algorithm is then proposed to find the structure similarity between two tree schemas. The proposed schema matching approach was evaluated by testing on several benchmarks of real schemas and comparing with other methods. The experimental results show that the proposed lexical decision tree substantially improves the linguistic similarity matching effectively and efficiently. The proposed SYM algorithm also performs high effectiveness on 1-1 schema matching.

No abstract

No abstract

Developing pervasive context-aware systems to construct smart space applications has attracted much attention from researchers in recent decade. Although many different kinds of context-aware computing paradigms were built of late years, it is still a challenge for researchers to extend an existing system to different application domains and interoperate with other service systems due to heterogeneity among systems This paper proposes a generic context interpreter to overcome the dependency between context and hardware devices. The proposed generic context interpreter contains two modules: the context interpreter generator and the generic interpreter. The context interpreter generator imports sensor data from sensor devices as an XML schema and produces interpretation scripts instead of interpretation widgets. The generic interpreter generates the semantic context for context-aware applications. A context editor is also designed by employing schema matching algorithms for supporting context mapping between devices and context model.