从Protege本体中获取JenaModel

最新推荐文章于 2021-12-21 22:46:18 发布

iteye_10155

最新推荐文章于 2021-12-21 22:46:18 发布

阅读量680

点赞数

分类专栏：本体之Protege 文章标签： MySQL 数据结构 JDBC

本体之Protege 专栏收录该内容

4 篇文章

订阅专栏

本文介绍如何从Protege中导出本体至Jena模型，并利用Jena推理机进行推理过程。提供了从本地项目及数据库导出模型的具体实现代码。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

因为项目要用到Protege中构建的本体，这些本体又在Protege中编辑，所以存储结构是使用其内部的结构。要想使用Jena推理机，必须要将Protege中的KnowledgeBase导出为Jena内兼容的结构。庆幸的是，ProtegeAPI提供的转换接口。

（1）从本地项目中导出Jena Model，并打印所有的Triple，程序如下：

	public static void getLocalModel() {
		Collection errors = new ArrayList();
		Project p = new Project("d:\\human.pprj", errors);
		OWLModel owlmodel = (OWLModel) p.getKnowledgeBase();
		Iterator<TripleStore> listtriplestore = owlmodel.getTripleStoreModel()
				.listUserTripleStores();
		while (listtriplestore.hasNext()) {
			TripleStore ts = listtriplestore.next();
			System.out.println(ts.getName().toString());
			Iterator<Triple> itriple = ts.listTriples();
			while (itriple.hasNext()) {
				Triple triple = itriple.next();
				System.out.println("triple:" + triple.getSubject() + ","
						+ triple.getPredicate() + "," + triple.getObject());
			}
		}
		owlmodel.getOntModel().write(System.out);
	}

（2）从数据库导出Jena Model，并且进行推理，程序如下：

	/**
	 * 从Protege本体数据库中导出Jena本体，并且进行推理
	 */
	public static void connectTest() {
		String driver = "com.mysql.jdbc.Driver";
		String url = "jdbc:mysql://IPaddress:3306/protege?useUnicode=true&characterEncoding=UTF-8";
		String table = "test";
		String user = "root";
		String password = "root";

		OWLDatabaseKnowledgeBaseFactory factory = new OWLDatabaseKnowledgeBaseFactory();
		Collection errors = new ArrayList();
		Project prj = Project.createBuildProject(factory, errors);

		try {
			OWLDatabaseKnowledgeBaseFactory.setSources(prj.getSources(),
					driver, url, table, user, password);
			prj.createDomainKnowledgeBase(factory, errors, true);
			OWLModel owlmodel = (OWLModel) prj.getKnowledgeBase();
			Model ont = owlmodel.getOntModel();
			ont.write(new FileOutputStream("onto/human_pro.rdf"));
			List<Rule> rule = Rule.rulesFromURL("rules/humantest.rules");
			Reasoner reasoner = new GenericRuleReasoner(rule);
			InfModel inf = ModelFactory.createInfModel(reasoner, ont);
			inf.difference(ont).write(
					new FileOutputStream("onto/human_infer_pro.rdf"));

		} catch (Exception e) {
			e.printStackTrace();
		} finally {
			prj.dispose();
		}
	}