Context and motivation: Multiple-viewed requirements modeling method describes the system to-be from different perspectives. Some requirements models are then specified in various UML diagrams. Question/problem: Managing those models can be tedious and error-prone, since a lot of CASE tools provide poor support for reasoning and consistency checking. Principal ideas/results: Ontology is a formal notation for describing concepts and their relations in a domain. Since software requirements are a kind of knowledge, we propose to adopt a knowledge engineering approach for managing the consistency of requirements models. In this paper, an ontology for three most commonly used UML diagrams is developed in Web Ontology Language (OWL). The transformation of UML class, sequence and state diagrams to OWL knowledge base is presented. Owing to the underlying logical reasoning capability of OWL, a semantic query language, SPARQL (SPARQL Protocol and RDF Query Language), is used to query the knowledge base for consistency checking. Contribution: This paper introduces a semantic web-based knowledge engineering approach to represent and manage software requirements knowledge in OWL. By experimenting with a concrete software system, we demonstrate the feasibility and applicability of this knowledge approach.

中文翻译：

---

利用 SPARQL 查询进行 UML 一致性检查

背景和动机：多视角需求建模方法从不同的角度描述了系统。然后在各种 UML 图中指定一些需求模型。问题/问题：管理这些模型可能很乏味且容易出错，因为许多 CASE 工具对推理和一致性检查的支持很差。主要想法/结果: Ontology 是一种形式化的符号，用于描述领域中的概念及其关系。由于软件需求是一种知识，我们建议采用知识工程方法来管理需求模型的一致性。在本文中，三个最常用的 UML 图的本体是用 Web 本体语言 (OWL) 开发的。介绍了UML类图、序列图和状态图到OWL知识库的转换。由于 OWL 的底层逻辑推理能力，语义查询语言 SPARQL（SPARQL 协议和 RDF 查询语言）用于查询知识库以进行一致性检查。贡献：本文介绍了一种基于语义网络的知识工程方法来表示和管理 OWL 中的软件需求知识。通过试验一个具体的软件系统，我们证明了这种知识方法的可行性和适用性。