Meta Search Based Vertical Search System Design

Comparing with the traditional full-text search engine， the vertical search enginecan-provide more professional and effectively information．The users can avoid filtering usableand professional information from mass data when they use the vertical search engine.This isresearch focus of search engine．At present，the vertical search engine adopted the systemstructure which is similar with the full-text search engine．It had hill level in the professionalassociation degree．Conforming to the universal full-text search engine，it is also confronted withthe problem of low-recall ratio and high dissipation of web resource．This article provides thesystem structure of the vertical search engine which based on rectal-search to solve theseproblems，the level of the professional association degree is declined,but the recall ratio isimproved．The experiment proved that this system can meet the needs of professional usersadequately．Works achieved in this paper as follow:1．At present, the recall ratio of the vertical engine is low. Because of Meta-search enginehad highest recall ratio,we design a vertical search engine which adopts the method of theinformation collection which is based on the Meta-search engine.The system adds the function ofinformation collection and analysis to adapt to the demand of the vertical search engine.2．Information collection is the foundation of the search engine. In allusion to the low-recallratio and inefficacious information, this article provides a method which based on the browsingtimes of the webpage. This method collects a mass of high attention information which based onthe Meta-search engine results. On one sides, it can increase the recall ratio of the search engine.On the other sides,I it also can improve the professional association degree of the information.3．Information search is the core of the search engine. This article imports data mining toanalysis the collected information and it reduces correlative rules between words and key wordsform the well-pleased information. We provide the concept of the adelomorphic key wordswhich used in the information rearch. Experment proves that the search which added theadeomorphic key words can improve the professional association degress of the resulte.4．The users concern the anterior results of the search engine, so the sorting of the resues isa important point which we should attach importance to. Now, the sorting of the results ofmeta-search engine use a little of information, it cannot make sure the result of search enginecorrelated with the specialty of the vertical search engine. This article provides an array methodwhich is suit for this system and improved the position sorting algorithm, it adds the professionaldegrees of association of the results.This article procides a solution of the vertical search engine which is based on themeta-search engine, it provides a new thought of the vertical engine designing.