Search result optimization using user search history

Search service which offers same result list to users using same query cannot satisfy user preference or burdens users with additional effort to find their interesting data. Listing up user preference data in first page of search result is the major factor for search performance. It can be a good work that search service offers suitable result to each user when several users having each interest area or major field inputted the same query. So we propose a method to personalize the search result using user search history data. And we make a target at service academic paper data having some classification codes such as the DDC (Dewey Decimal Classification) code. The search result personalization proceeds in two steps. user preference areas are first detected by analyzing search history data which are user query, search result per query, and records about user´s detail view paging. And then boosting algorithm using pre-detected user preference are makes personalized search result. This rule lists up the good data to each user even if several users having different interesting area input same query.