Font Size: a A A

Research On Application Of Text Classification For Mobile Internet

Posted on:2017-03-30Degree:MasterType:Thesis
Country:ChinaCandidate:J ChengFull Text:PDF
GTID:2348330491951549Subject:Information security
Abstract/Summary:PDF Full Text Request
With the rapid development of mobile Internet technology, mobile networks gradually become the most important platform for information exchange network. Mobile Internet data with unprecedented growth rate has brought enormous difficulties in data analysis and network management. Data classification technology research for mobile Internet has important practical significance. Text classification which is widely recognized as an effective technology has been widely used in the Internet field. Applied in the data classification and management, text classification can get rid of the shackles of mobile terminals, platforms and applications with automation efficient, cross-platform and many other advantages.The paper carries on a series of research work for the classification method of mobile Internet data. According to the characteristics of mobile Internet data, the paper puts forward a new classification model and experimental verification, and ultimately builds Mobile Internet Data Analysis System based on the improved technology. The main work can be summarized as follows.(1) The paper analyses on text representation, weight computation, feature selection and classification, which builds on a solid theoretical foundation for the study of the later work. Summarizes the network traffic classification method and its characteristic.(2) According to the characteristics of the actual data of the mobile Internet, combined with the advantages of different classification methods, the paper proposes a hierarchical analysis model, which classified the mobile HTTP data into two levels. In the coarse-grained classification, text classification is used to classify the HTTP data of mobile Internet. We improved the key technology and experimentally verified its validity. In fine-grained classification, according to different classification data, feature string matching method is used to classify the mobile data,(3) According to the research and improvement of classification technology, this paper designs the Mobile Internet data Analysis System, implements the function of real-time data acquisition, protocol analysis, data classification and application identification, provides the basis of further analysis of mobile Internet user behavior and network features.
Keywords/Search Tags:Text Classification, Mobile Interact, Data Analysis, Feature Extraction, Application Identification
PDF Full Text Request
Related items