SemFilter: A Simple and Efficient Semantic Filtering of XML Messages

Until now, most researches for XML message filtering have focused on the efficient and scalable path matching of a large number of XPath queries over a continuous XML data stream. However, we think there is another key problem of the semantic interoperability of subscribed XPath queries against heterogeneous data sources. Recent studies only assume that the filtering system maintains a single global schema covering all data sources, or the data sources follow the defined global schema. However, beyond the simple assumption, the data sources can have various different element tag structures and names, although related with a similar subject. Therefore, the path expression of a subscribed query against a specific data source cannot be matched with other data sources that have semantically the same information. In order to resolve this problem, a query translation method rewriting the subscribed path expression into one corresponding to each data source is required. In this paper, we introduce the problem of semantic interoperability in XML message filtering and suggest a novel query translation method.