Research On The Methods Of Micro-blog Centralization For Online-Advertising Targeting

With the unique method of dissemination, micro-blog has become the main form ofinternet information service, which has changed the way people get and exchangeinformation. The useful data hidden in the micro-blogs have provide natural platform fortargeted advertising of internet. How to extract important information from thefragmentized and socialized micro-blog has become an urgent demand for the internetservice provider.The background of this thesis is websites which provide micro-blog services. Theobject is micro-blog. The content is researching on micro-blog centralization methods. Thepurpose is to solve problems mentioned above. This thesis first proposes the purpose andmeaning of this research, then summarizes the study of domestic and foreign scholars inmicro-blog centralization and targeted advertising fields, and proposes the classification ofmethods of micro-blog centralization. After that, the thesis puts forward the concept oftargeted advertising specially used in micro-blog platform, and proposed classification ofadvertisement targeting. Based on this, the micro-blog centralization model is proposed.Then the process of content centralization and user centralization is designed by usingmethods like material sieving, textual opinion mining and user feature analysis. The wayto combine content-oriented targeting and behavior-oriented targeting can find audiencefor advertising more precisely. Finally, an example based on the model is built and theresults are analyzed to verify the theoretical results.