Service Materialization

Search Computing queries typically obtain data by invoking remote Web services. In certain cases, this may be unfeasible, e.g. if the service is too slow or if its access pattern requires data in input which is not available at query execution time. A possible solution in that cases is to materialize the service data (i.e. download it) in a local database to be efficiently queried at runtime. The goal of the proposed project/thesis is to realize a flexible and efficient Web service materialization solution for Search Computing, that:

• permits to efficiently materialize selected Web services and keep up-to-date materialized data;
• requires minimal configuration;
• permits to identify which Web services will benefit more from materialization, based on available service statistics and an estimate of materialized data source performances.

Previous knowledge of Web service and database technologies is useful. The work is part of an existing framework that hides the complexities of dealing with the invocation of heterogeneous Web services. A previous materialization solution is also available and can be used as a starting point for the work.