Analysis of University Public Opinion and the Mechanism of Maintenance Stability in the Media Age

An analysis mechanism of university public opinion in self-media age is put forward based on parallel KNN clustering algorithm to increase accuracy and efficiency of university public opinion analysis for the problems that university public opinion analysis has been increasingly complicated in self-media age. First, the characteristic extraction and similarity calculation for university public opinion are rapidly clustered through topics on university public opinions; then, it is essential to carry out parallel MapReduce-KNN clustering algorithm design and give algorithmic computation complexity analysis; finally, the advantages of algorithm put forward on clustering quality and clustering are verified through experimental comparison.