Semantic Keyword Detection of Web Document

摘要

This paper deals with a method for semantic keyword detection of web documents. While the document consists of many kinds of part-of-speech, we just consider noun words. We propose a method for measure of relation value that represents relationship between words occurred in the document using WordNet. We calculate concept weight that measures semantic importance of words by using term frequency and relation value together. According to the concept weight, we detect the 10 keywords of document in order. Through experiment, we show that our method is possible to extract main keyword semantically from document.