Research On Search Engine Of Mobile Intelligent Terminals Based On Cloud Computing

Search engine, as one of the most popular application of Internet technologies,has been a search focus of academic and information industry for recent years. It isan important way for people to access information. Therefore, how to retrieve usefulinformation from the mass of information quickly, accurately and control the loadbalancing of node for network resource are the important standards to measure thequality of the search engine.With the expansion of information data, the current search engine is facingsome serious problems, such as limited storage space and computing power. What’smore, the explosive growing number of users and the high concurrent request arealso posing a challenge for search engine. So, how to improve the storing andprocessing power to fulfilling the needs of users is of great significance. In thewhole research process, the author analysed the key factors which affect thefunction of the search engine. Then, the author made deeply study of informationretrieval and data storage. The main work of the paper is as follows.Firstly, the paper discussed the shortcomings and technical bottlenecks of thecurrent traditional search engine. Then, in the understanding of search enginefeatures and technical requirement, it improved the system by means of the cloudcomputing architecture. With the open source cloud computing platform of Hadoop,the paper designed and implemented the search engine of the various modules andfunctions.Secondly，For solving the problem of low searching efficiency, the paperproposed a new cloud classification model of index data storage. By the parallelcomputing power of Map/Reduce, the accuracy of information retrieval and recallrate were all improved well. In addition, during the study of metadata managing mechanism (HDFS), thepaper optimized the load balancing algorithm by combining the current loadbalancing algorithms and the cyclical feedback mechanisms of data management.The improved system not only balanced the load between server nodes and datanodes, but also saved network bandwidth during data transmission.Finally, the paper test the whole system performance in the laboratoryenvironment, demonstrate the superiority of the system by analysis of experimentaldata.