Multi-tool Methodology for Converting Text into RDF Triples

Abstract

There is no doubt that Internet becomes one of the most important sources of information. At the same time the amount of information stored on the web and available for users becomes enormous. In order to make this information more accessible and create prospects for software to process it automatically, a different format of storing information has been proposed by World Wide Web Consortium – it is called Resource Description Framework (RDF). This format can be described as a triple: subject-property-object. Application of RDF leads to creation of a highly interconnected network of nodes containing pieces of information. RDF could change how information is stored and processed on the web. However, one of the most common formats of representing information on the web is and will be a simple text. A textual format is the most natural way of representing information used by individuals. Therefore, in this thesis, we focus on the task of translating text into RDF. The proposed approach is based on a combination of well-known Natural Language Processing tools: parsers, with a web-based tool for disambiguation, and our own algorithms for combining the results obtained from these tools and converting them into RDF triples.