TITLE:
Research on Network Group Event Recognition Based on Neural Network
AUTHORS:
Haiying Ma
KEYWORDS:
Sina Weibo, Network Group Events, Hot Event, Pattern Recognition Model
JOURNAL NAME:
Open Journal of Social Sciences,
Vol.4 No.12,
December
27,
2016
ABSTRACT: Sina Weibo is one of China’s most popular SNS platforms, which has a large number of user groups. Nowadays, the public opinion on Sina Weibo has great influence and significantly affects the social reality. This paper defines the network group events from aspects of the nature of the incident and influence and establishes a neural network of pattern recognition model that is used to fit the model. The model can recognize the network group events through continuous three days’ various topics growth rate. By programing a crawler, we get the data of 23 hot events on Sina Weibo from 2011 to 2015 for neural network training. The training results of pattern recognition model show that the accuracy rate of the judgment of the Internet Group events and the non-network group events is 72.7% and 75% respectively. The overall accuracy was 73.9%.