This Ph.D. thesis deals with the creation of ontologies using machine learning algorithms. Ontologies are artefacts that are being developed to facilitate knowledge sharing, exchange and re-use. Ontologies have become a hot research topic in many research communities of Artificial Intelligence. They promise shared, comprehensive and machine readable and processable knowledge through a controlled vocabulary, which describes domain knowledge, and common domain theories, which denote the structure and semantics of knowledge. Although, ontologies have proved to be beneficial for many application areas, such as natural language understanding and information extraction, their creation is a time-consuming process which requires an interdisciplinary approach and lot of mental effort. Nowadays, the need of ontology building and maintenance in a robust way indicates research on the emerging field of Ontology Learning. Ontology Learning is defined as a sub-field of Artificial Intelligence that aims to alleviate, as much as possible, the intervention of human onto the tedious and time-consuming process of ontology building and maintenance by using machine learning techniques.The aim of this thesis is the specification of a method for the population of an ontology using
machine learning and natural language techniques from web-pages. We think that this
method will be beneficial for the creation of application and/or domain ontologies from the
appropriate web-pages, as the WWW is the largest information repository ever, and for the
semantic annotation of the content of web-pages, as the conversion of the World Wide Web
into a Semantic Web is not a dream any more.