Research of algorithm for automatic opinion leader detection in BBS
Deren Chen · Systems Engineering - Theory & Practice · 2011
Detection ofopinion leadersautomatically in BBS is significant for us to grasp and guide the public opinion in time.Consequently,we proposed an algorithm which can detect theopinion leadersin BBS automatically.After analyzing the essential attributes ofopinion leaders,a X-means iteration clustering filter model based on the Bayesian information gain maximization formula was proposed to find out the opinion leader feature points in the feature space.The experiment shows that the algorithm based on the model has excellent performance on precision and recall,hence the method can monitor the public opinion on the network effectively.