Design And Development Of Weibo Content Monitoring And Anti-Spam System Based On LNMP

Sina Weibo is a social platform for public online leisure information sharing and communication which was found by Sina in 2009.It has greatly changed people's way of information fetching and communication since its coming out,and has become one of the biggest social platforms among Chinese communities.According to the Weibo users development official data report in 2016,the monthly active users of Weibo have come to 313 millions in 2016,and the amount of tweeting,retweeting and commenting got a year-on-year growth by more than 50 percents.The explosive contents also bring about the illegal information blooding such as violent,erotic and politically sensitive contents,and ads,harassing information in addition.The apartment of Weibo content monitoring faces severe challenges.To improve the efficiency of Weibo monitoring work,a Weibo content monitoring and anti-spam web platform was designed and developed on the base of existing Weibo monitoring system,which realized the combination of manual audits and automatic disposal system.The new system is a web platform connected to the existing automatic Weibo monitoring system in order to manage those automatic subsystem well.The system has accesses to the total users' and blogs' data and offers a more efficient platform for manual reviewers by dividing modules of users management,content audit,anti-spam management,keyword management and so on.The system also has power management module for role definition.Different users accessing the system will be authorized with different power to ensure the security of the system.The system adopted the classic MVC design modeling to separate the view,model and data interaction well.It was developed in open source LNMP framework.The key-value cache database Memcached was used to share the heavy pressure of MySQL,and the asynchronous message queue Memcache Q was also used to deal with those time consuming tasks.Function and pressure tests were performed to verify the availability and performance of the system.The test results show that the developed system meets the design intention to increase the efficiency of Weibo content monitoring.And the LNMP framework was validated with high concurrency,low resource occupation and high stability.The splitting strategies on Weibo content storage was validated with high efficiency as well.