Managing scientific computations with a Java-based schema extension facility

Scientists using several computer programs to produce and/or analyze computational data face several challenges. They must organize the invocation of the programs. Also, they must keep track of the voluminous data being generated and/or analyzed. In addition, they must record the steps involved in the computational process for later examination.; We have produced a Java-based Schema Extension Facility (JSEF) as an aid to scientists involved in the process of computational discovery. We use Java's object-oriented data model as a basis for classifying computational entities. Also, JSEF's relationship mechanism allows us to create relationships between classes. In fact, JSEF allows a series of associated programs to be classified as a single entity called an analysis strategy.; In the computational process, each entity is referenced by a JSEF object. In order to drive the computation, a method is called on a JSEF object, which invokes the necessary processing. JSEF objects are stored in a database upon creation. Thus, JSEF reduces the tasks of invoking and tracking a computation and the entities involved to a single method call and subsequent querying of the database and examining the state of the JSEF objects in the query result.