Research And Implement Of Micro-blog Interest Found And Pushing System

As one of the most widely used applications in Web2.0era, micro-blog has thecharacteristics of dispersion, diversity, timeliness, and the huge amount of information andso on. In order to get the user’s interest from large, distributed information, and to push itto users in real time, this paper has done some research on micro-blog interest found andautomatic pushing.Based on the analysis of the features of micro-blog information and the user behavior,and combined with the domestic and international technologies associated with interestfound and micro-blog interaction, this paper focuses on the technologies of informationcollection, Chinese word segmentation and interest found for Chinese micro-blog system.The main work of this paper is as follows:Based on the analysis of features of micro-blog releasing, a java scheduling techniqueis proposed to collect micro-blog information by using related interfaces periodically;Through the studies of domestic and international Chinese word segmentation systems, aChinese word segmentation program based on Hidden Markov Model and user dictionaryis used to divide sentence, a stop word processing method is used to reduce theinformation redundancy; Text classification and subject dictionary matching is usedtogether to find user’s interest after the research of micro-blog user behavior; A monitor isproposed to be used in micro-blog pushing timely based on the characteristic of timelinessof micro-blog system.This paper has done detailed analysis of the results on the background of Weibosystem. The experimental results show that, the proposed system has high accuracy andcould be used to find user’s interest well, which indicates that it has some value inengineering.