Research Of Microblog Retrieval Based On The Thought Of Federated Search

With the advent of Web2.0 era,there are more and more applications in the Internet,and the user’s participation in the internet is gradually improving,the internet with more people is also marching towards socialized.Microblog service is the most typical application in the socialized internet,which has attracted more and more users with its characteristics of simple content and easy release.With the inc reasing number of users in microblog,the content that users generated in the microblog platform is also increasing.However,the search for the content of microblog is still using the traditional centralized search mode,which brought some problems to the users.First of all,due to the huge amount of microblog data,searching all content directly will waste a lot of time and reduce the users’ search experience;secondly,there are too many themes on microblog which may cause a reduced accuracy of the centralized search;finally,centralized search can only use one search model,but federated search can adopt different models for different search data sets and can be more flexible.Federated Search is an important branch of information retrieval,which can search different data sets,and solve the problem of low effic iency and accuracy.First,the Federated Search will determine the relevance of e ach data set and query words,and then send the query words to a relatively large data set for retrieval.Because the query data sets are relative to each other,the search results are more accurate than the centralized ones,and the problem that the data set is too large to be retrieved effectively is solved at the same time.Based on the advantages of Federated Search,this paper proposes a microblog search technology based on the idea of Federated S earch.The technology will apply the concept of Federated Search to the microblog search field,while taking the specificity of microblog text and the authority grades of different authors into account,so that the calculation of the document ranking score become more accurate.The experimental results on real microblog data sets show that the proposed method in this paper can improve the accuracy of twitter search.The main work of this paper is as follows:(1)Develop a microblog search framework based on Federated Search.This paper focuses on the use of Federated Search technology to search for microblog data.To this end,the first step is to establish a federated microblog search data set and generate the description for each data set.T hen using the data set selection method and according to the established data set description to calculate the matching score for each data set and query words.The data sets are sorted according to correlation,select several most correlated sets next.And then the query words will be sent to the selected data sets for searching.Finally,the results of different data sets will form a single list of search results,and the list will be se nt back to the users.(2)Present a new result merging algorithm to merging the results contain the authority of microblog authors.This paper,based on the characteristic of microblog and the previous studies,proposes a method of combining the results of the authority of microblog authors.The experimental results show that the proposed method can effectively improve the accuracy of the search results compared with the previous method.(3)Design a microblog search system based on the idea of federated sear ch.On the basis of the first two chapters,a basic microblog search system based on the idea of federated search is designed.This system mainly includes four functional modules,such as the establishment of microblog index,the management of microblog text,general search and Federated Search.Finally,the paper also shows the system.