Reduced impedance mismatch in speech database access

Database access can be seen as a series of communicating systems each contributing to the overall efficiency of the query. This paper investigates the effect that impedance mismatch has on speech database access: the efficiency of database transactions between the corpus and the querying application(s). A system where impedance mismatch has been reduced through an object-oriented paradigm is presented and several query examples are shown. The paper suggests that for applications requiring high rates of database access for data with many expressed relationships a tightly coupled object-oriented system extending from the corpus to the end-user application is required.