The Research And Analysis On SNA Of Telecommunications On Hadoop Platform

The huge social network of telecom contains a lot of potential values. With the continuousgrowth of structured and unstructured datas, the efficiency of the huge datas processed bytelecommunications industry getting lower and lower, because it is difficult to dig out from the bigdatas. Development of cloud computing to effectively solve this problem, the open source cloudplatform, Hadoop cloud platform has been widely applied to various fields to improve the ability byprocessing data. Telecom’s existing data warehouse will be migrated to the platform of Hadoop. Bymaking full use of HDFS and MapReduce model, mining the telecom social networks, thetelecommunications industry can enhance the market competitiveness in a cost-controlled way.This paper first introduces the basic theoretical knowledge of Hadoop cloud platform, thenintroduces the telecommunications data warehouse and the analysis of social research status oftelecommunications networks. This paper primarily through two parts to improve mining efficiencyof telecommunications circles: first through analysising the popularity of documents to improve thedata backup mechanism dynamicly, to enhance the ETL (extract, transform and load) efficiency andthus decrease the preparation time during data mining; secondly taking advantage of MapReduceframwork, improving existing graph algorithms, combining the directed graph with the undirectedgraph to find the frequent exchanges telecom circles accurately.This article improves the efficiency of the mining of telecommunication social network.Telecommunications industry by mining the communication circlesand combine with manydifferent data mining algorithms can draw many potential business informations, with can help tomanage markets and customers efficiently. Mining personal social circles, the core task is focusingthe concern customers, making precise marketing, so that to enhance the customer experience, tomeet customer needs in real time.