A new filtering-based query processing: improving semantic caching efficiency in mediation systems

Due to the exponential increase in information requirements on the one hand1, and in the number of heterogonous and distributed sources on the other. Several integration approaches, including information mediation have been suggested. The main purpose of information mediators is to let end users consume data sets that can be issued from a multitude of heterogeneous sources without paying attention to their specific characteristics. Nevertheless, involved sources might be geographically dispersed and therefore inaccessible from time to time. It is in this context that numerous studies on semantic caching have been carried out to speed up mediation process. To this end, previous user queries are cached so their results can be reused to answer future queries. However, existing semantic caching approaches might lead to extra costs in case the cache does not contain enough results. In this paper, we describe a new semantic caching mechanism that makes it possible to both reduce response delays and to avoid extra costs generated by the usage of caching.