The Micro Blogging Precision Advertising System Architecture Design Based On User Rank And Users’ Interest

The popularity of the Internet and Web 2.0 promoted the rapid development of micro-Blogging, and provided the prerequisite of micro-blogging precision advertising. The advertising system aims at placing the advertisement to its target audience who is really interested in it going, thus the user’s interest mining is the key point of the system. If we serve the ads solely based on users’registration information, the ads may not perfectly match its target audience because of incomplete information, update-slow information and false information. To compensate for this deficiency, this paper presents the new idea to mining users interests by means of the basic information, the relationship circle (follower & following), and the topic participating in. On the other hand, the current evaluation of advertising effectiveness is based on the number of forwarding and commenting tweets, but the followers number and the authority of each person who forwarding and commenting the tweets is different, so as to the forwarding and commenting effect of them. As a result, if we barely use the number of comments and forwards to evaluate the tweet advertisement effectiveness, the result is flawed. To solve this problem, this paper presents the concept of User Rank to evaluate the influence of micro-blogging users, the thought of User Rank is similar to Page Rank, and then we add User Rank to the assessment of ads. effectiveness.This study started in two ways:1. The design of precision advertising system based on user interest mining,2. The design user’s influence (User Rank) index. For the first part of the work, the authors gives advertising accurate system design and describes the function of each module by learning the data mining techniques, LDA model and Chinese word segmentation. For the second part of the work, the authors,and design User Rank algorithm by studying the various factors that affect the user’s influence power and evaluate the results of algorithm by processing Sina micro-blogging’s data, and also the author compared the result with other two algorithm.