A knowledge-based rapid prototyping system (KBS) is developed to assist system analysts and users to create consistent and complete requirement specifications and software prototypes. In the prototyping process, a requirement model, called RM, is used to model data intensive applications. This model is based on six modeling concepts which can capture more static properties and dynamic behavior and facilitate the generation of software prototypes. The consistency and completeness verifications provided by this system are formally defined by means of integrity constraints, inference rules, and algorithms. The integrity constraints describe the semantics of four kinds of relationships among objects, such asinstance-of, part-of, is-a, and disjointness. The inference rules are used to deduce implicit relationship embedded in the acquired requirements. These relationships and inference rules are mapped into a digraph. Thus, the algorithms for these verifications can be transformed into the searching problems in Graph Theory.