Improving the UML consistency using Text Semantic Similarity approach

It is an important and stimulating issue to discover the inconsistency and incompleteness of the large software system through the UML diagrams. However, the use of temporal logic and model checking techniques to address the problem of UML consistency for developing software systems has received much attention. It is still not applicable and hard mission to specify the different consistency issues of UML. In this paper, we address this problem. We introduce a novel approach for checking the consistencies and the incompleteness of large software systems that are designed using the different UML diagrams. This novel approach uses one of the recent natural language processing techniques, called Text Semantic Similarity (or Textual Entailment), to improve the consistency and the completeness among various UML diagrams that model the same system.