| Abstract [eng] |
Worldwide technology advances aimed at creating a user friendly semantic network. Also facilitate human-software interaction, ie to create conditions for computer to understand natural human language. One of the steps in order to achieve this goal is the automatic semantic annotation. Currently, most of the information is on the website. With such a large amount of information, it becomes difficult to find exactly what you need. The search in semantically annotated texts can obtain more accurate results than using a simple web search. However, in order to semantically annotate the text, specialized dictionaries of domains and ontologies have to be created, suitable annotation methods chosen, semantic annotation rules described and tools for that created. It's especially important for Lithuanian language to develop tools of semantic annotation of text because it is behind in this area compared to other languages. The main purpose of this work is to create a methodology to recognize sport events in the Lithuanian text and provide an ontology which allows to annotate such text. According described methodology it is possible to recognize achievements defeats, results and participants of event. Created methodology and described sport events rules can be used to create an automatic semantic annotation tool. Ontology was created in OWL language, using ontology development tool \" Protégé \". Because tool is free and chosen language is widely used, everyone can use and modify this ontology. Sporting events ontology can be supplemented new classes or other ontologies. SPARQL queries were developed, which are used to perform searches in sporting events ontology. |
