Web sites are usually semi-structured and§therefore very hard to query for information.§Traditionally have keyword searches and word§comparison been the way the§search engines do the searching. They have no§understanding of the real meaning of words and their§relations in a context. More advanced extraction§processes can be used to gain§better results. Based on ontology we can define§instances of words with mutual relations in§information sources and use this as a guidance of§what we want to extract. The extraction itself is§done by natural language processing and...