Abstract: Vertical search engine provides such services for people, which treats professional knowledge as topics and searches the information related highly to the same topic. The present topic relativity judging algorithms cannot match the speed of information growth. For the match, a topic relativity judging model based on semantic information is provided. The experiment result shows that the performance of the algorithm significantly improves the ability of judging topics, and this method is of great practical value.

Key words: Internet, vertical search engine, topic relativity judging, semantic information