Automated Test Code Generation Based on Formalized Natural Language Business Rules
The paper addresses two fundamental problems in requirements engineering. First, the conflict between understandability for non-programmers and a semantically well founded representation of business rules. Second, the verification of productive code against business rules in requirements documents. As a solution, a language to specify business rules that are close to natural language and at the same time formal enough to be processed by computers is introduced. For more domain specific expressiveness, the language framework permits customizing basic language statements, so called atomic formulas. Each atomic formula has a precise semantics by means of predicate and Interval Temporal Logic.