Research On General Frame Model Of Vertical Search System Based On Meta Search

With the growing amount of information on the Internet,and the more clear data classification,the general search engine is not suitable for different background and professional users to find the information they need.Therefore a professional search engine is needed so that it can retrieve a specific domain knowledge.Vertical search engine improves the accuracy of the search results by focus on a specific domain,crowd,or demand,and provides users with highly targeted and specialized content.However,there are many problems as with the full-text search engine: lower recall rate,higher network resource consumption and so on.According to some statistics,within the scope of the entire network,a considerable part of the page can't be directly indexed by the search engine.Even for the largest search engine---Google,it can only index thirty to forty percent web pages.The solution to this in this paper is to establish the concrete structure of the vertical search engine on the Meta search system,and to improve the recall rate of the system by means of multi-source data.Recently there are many studies on vertical search,and many vertical search tools for different areas,but these vertical search tools are not reproducible and reusable.At the same time,they are independent of each other.Also lack of a common framework that can support quickly build a new vertical search system.The main work of this paper is designed a generic Holonic retrieval model for vertical searching system based on Meta search technology,which supports the rapid construction of new vertical search system in other domains.And take the academic field as an example,to implement an academic Meta search tool.The core contributions of this article are as follows: 1.by analyzing the commonness of existing vertical search systems,a general framework model of vertical system is given.At the same time,based on the system structure of the Meta search system and the model of Holon,the framework model proposed before has been optimized,then a generalize vertical search system framework model based on Holon is designed to support the rapid construction of the vertical search system for a new domain.2.by analyzing the system objectives for the academic system,combining the general vertical search system framework model given in this paper,we designed an academic search system's framework based on Agent,and an academic search tool has been implemented(IM Scholar Search).3.through the experimental deployment and case description of the academic search tool,on the one hand,it is used to verify the validity of the generic Holonic retrieval model for vertical searching system proposed in this paper.On the other hand,it is mainly used to verify the usability of the academic Meta search tool realized based on the above structure,and ensure to meet the general needs of the user.Through the experimental verification,the academic Meta search too implement in this paper basically satisfies the general requirements of the users,which not only demonstrates the usability of the tool,but also validates the support of the framework proposed in this paper for the new domain vertical system,and validates the design effectiveness of the proposed framework.