Research On The Incremental Maintenance Of Materialized Views Based Enterprise Vertical Search Optimization

Improving and optimizing the technology of vertical search engine have beenconcerned all the time in the information industry. However, for a long time the research ofthis field is more concentrated in the network data crawling, retrieval efficiency improving,personalized services and so on. Even if in the enterprise level, the related research is moreconcentrated in the processing of web data, it seldom has the vertical depth of excavationagainst the knowledge and information of the single enterprise. This leads that the overallquality of the returning results of the enterprise vertical search is not high. How to improvethe service performance of the enterprise vertical search engine to provide more valuabledecision support information for the work, which has been become the focus in academiccircles and business community. This article will base on the practical needs, and in-depthanalyze the frame of enterprise vertical search engine, then form the two aspects of thesystem retrieves efficiency and personalized service to start. Among it in, it will stresses inimproving and optimizing on algorithm of incremental maintenance of materialized viewsand presenting of search results. Finally, combines a ship-supporting enterprise to verifyand evaluate what was introduced above.Firstly, contrary to the business search physical layer of the vertical search engine inenterprise and create view and with the form of materialization to be stored in the datawarehouse, use “cache” mechanism to improve the efficiency of the system retrieves.Because of the materialized view stored content has static characteristic, it needs theachievement of updating and maintenance of the materialized view in focus point. Thisarticle introduces the concept of semantic constraints and improves&optimizes onalgorithm of incremental maintenance of materialized views and search results, obviouslyimprove and enhance the retrieval efficiency of the system.Then, achieve the presenting of retrieval results of presentation layer by the way ofmeta-data-driven application, and combined with the characteristics of users demands tostore the personalized feature information by meta-data, use the second-stage view tomaintain and optimize the system performance.Lastly, unify the improving&optimization work above to the overall frame ofenterprise vertical search system to describe its specific application mechanism in detail,and use an application case in a ship-supporting enterprise to verify the correctness and effectiveness of the research in this article.