Apparatus for clustering same and similar product using vector space model and method thereof

摘要

PURPOSE: A device and method for clustering an identical and similar commodity using a vector document model are provided to cluster words which display commodities in a web page collected by a search word. CONSTITUTION: A web page collection unit(110) collects web pages corresponded to a query word in shopping mall sites(101). A single cluster forming unit(130) extracts words which explain the query word according to the web pages and forms a cluster. A vector conversion unit(140) forms a vector using the words according to web pages. A similarity calculation unit(150) calculates a similarity among vectors according to web pages. A cluster combining unit(160) combines clusters in accordance with the similarity. A character substitution unit(120) removes a special character included in the web pages and forms a character string comprising only general characters.