mahout框架下基于tf改善的vsm文本聚类研究控制工程专业论文.docxVIP

11 万方数据 AbstractAb Abstract Ab stract With the rapid development of the computer industry and Internet,the data which generated by the use of computer is also growing．Espechlly in recent years， the data generated by using of computer has shown an explosive growth．New challenges have come along with the huge amount of data．How to store such a huge data,how to explore from such a hu萨amount ofdata to discover useful inforrmtion that people interested in，are the problems which needed to be solved．Although big data presents a challenge to people，it also brings opportunities to people．If SOme useful information has been found fiom a hu萨data which belongs to a company,it will increase the company’S COmpetitiveness greatly．Today,more and more enterprises and institutions have joined to the ranks ofbig data research,especially on the aspects ofbig data text mining． Text clustering is a more popuhr technique，but also a major branch in the field of text mining．Unlike technology text chssificatiort,text clustering is an unsupervised learning method．Namely,text clustering does not requke prbr training throughout its procedure．In fact,Cluster analysis can be performed directly on the unknown data set,which is an advantage oftext clustering．Howeve r，the object ofthe traditional method of text clustering is stroll data,which is not suitable for big data text clustering．So big data text clustering must seek other methods．In academia and the business community，the commonly used method is to use Hadoop platform to text mining work for big data． Hadoop is the more popular and widely used distributed computing platform．As One of the sub projects of Apache Lucene，Hadoop technology has become increasingly mature．Sub project of its ecosystem is also increasing．The mahout project is One of the core projects in the Hadoop ecosystem,is a specialized tool for big hrge data mining．Compared with the vaditional text clustering technology,text clustering analysis by mahout in Hadoop clusters has