Using schema knowledge to optimize query evaluation, known as semantic query optimization (SQO), has generated promising results in XML query processing. In XML stream processing, we can use the schema constraints to expedite the traversal of the streams and to minimize memory consumption for holding the intermediate data during query evaluation. These are particularly critical for stream applications, which require real-time responses and both typically operate in limited main memory. However, as illustrated by the motivating scenarios below, these prior techniques assume that the XML schema is static and is available prior to the start of the query execution. As the scenario below highlights this assumption is unrealistic and thus may render existing techniques non practical.Assume that in a news publishing (or dissemination) scenario, the news server retrieves news from a large number of multiple sources, such as different reporter devices, different broadcast agencies, and government sources and disseminates such heterogeneous messages as an XML stream to subscribers.