| Recently,with the fast development of the Internet,more and more water-army areappearing on a lot of social media,such as BBS, mico-blog, blog, taobao(www.taobao.com)and so on. It is necessary for us to research and study water army behavior characteristics forthey are engaged in subject hype, event marketing and other harmful behavior, even interferewith the normal dissemination of public opinion, mislead public opinion direction, andendanger the safety of public opinion. In this paper, according to the extraction technology ofcomments based on functional semantic units and the related algorithm on a BBS for a periodof time of the release of the data; and extract the basic attributes of account, posts andcomments. According to the basis of artificial classification which was carried out on thenetwork account, we analyzes the distribution of water army posts and comments; and theirbehavior of the severely affect public’s opinions and conclusions on BBS. Posts andcomments are the result of water army behavior, thus taking posts comment time and responsetime as the research object, through the analysis of different periods and different time startingpoint distribution, the number of posts and the comments of statistics to find water armygroup behavior characteristics. The typical three peak hour period on distribution provideseffective evidence that water army group are collaborative.That also verify that thewater-army‘s behavioral characteristics is an organized group collaborative behavior.In addition, this paper also use the community discovery algorithm of the post ratealgorithm based on constructing the network comments diagram which is put forward by FanChunlong and other people. then we use Pajak software to build water-army’s organizationchart, so as to reproduce the behavior of water-army in the form of visual features andorganization mode. It can verify again that water army is a network organized group withhigh and deeply collaborative activities. |
PDF Full Text Request