Generation of Ontologies from Linked Data

2001年，Tim Berners-Lee引入了语义网这个术语，用来指代可以被认为是互联网的未来：一种机器可解释的内容的网络，可以由自动代理以有意义的方式进行处理。由于实现这一雄心勃勃的目标需要对相关领域知识进行解释和形式化，因此RDFS和OWL等本体语言已经成为明确知识规范的一种手段。然而，Tim Berners-Lee所设想的语义网的实现以及基于推理的智能应用的广泛使用仍然受到本体资源缺乏的阻碍。Web上存在的大量以RDF三元组形式存在的链接数据可以被认为是迈向语义网的道路上的重要一步，但它缺乏必要的表现力以及语义和句法的准确性，这是产生不明显结论的逻辑推理必不可少的要求。在这个项目中，我们将开发新的方法来获取结构化的知识表示，这些方法利用了Web of Data上的现有内容。通过将逻辑和统计关系学习应用于海量的关联数据，同时尽可能重用人工设计的形式本体，我们希望提供引导语义网实现的方法。该项目的一个重大贡献将是开发新的综合学习方法，这种方法可以在可伸缩性和稳健性方面满足链接开放数据的特殊要求。实验将证明我们的方法在不同领域和应用场景下的效率和有效性，其中最核心的是知识库的自动本体工程支持和调试任务。

In 2001, Tim Berners-Lee introduced the term Semantic Web in order to refer to what can be perceived as the future of the internet: a web of machine-interpretable content that can be processed by automatic agents in a meaningful way. Since achieving this ambitious goal requires both an explication and formalization of relevant domain knowledge, ontology languages such as RDFS and OWL have emerged as a means for unambiguous knowledge specification. However, the realization of the semantic web as envisioned by Tim Berners-Lee and even more the wide-spread use of intelligent, reasoning-based applications is still hampered by the lack of ontological resources. The vast amount of linked data in the form of RDF triples which is out there on the Web can be considered an important step forward on the way to the semantic web, but it lacks the necessary degree of expressivity as well as the semantic and syntactic accuracy which is an indispensable requirement for logical inference that yields non-obvious conclusions. In this project, we will develop new methods for the acquisition of structured knowledge representations which leverage existing contents on the Web of data. By applying logical and statistical relational learning to the vast amounts of linked data while reusing manually engineered formal ontologies whenever possible we hope to provide means for bootstrapping the realization of the semantic web. A significant contribution of the project will be the development of novel, integrated learning approaches, which can handle the particular requirements of Linked Open Data in terms of scalability and robustness. Experiments will demonstrate the efficiency and effectiveness of our methods when it comes to different domains and application scenarios the most central one being the task of automatic ontology engineering support and debugging of knowledge bases.