Abstract
Nowadays, the explosion of information is a majorsource of trouble for web searchers. Utilizing better and betterpage ranking mechanisms, web search engines do their best topresent information as relevant to users¡¦ queries as possible. Generally speaking, web search engines return lists of webpages related to query criterion as results. However, for endusers, this may not fit his/her expectations. It is highlypossible that a ¡§group¡¨ of web pages are needed to satisfy auser¡¦s need in one searching session. As a result, we proposea relation-oriented web querying technology. Instead ofperforming query on the basis of standalone pages, wepropose to perform query on the basis of groups of relatedpages. Besides, we would like to take advantage from crowdintelligence to improve search efficiency. The concept and theguideline of implementation are presented in this paper.