Title :
Burst events detection on micro-blogging
Author :
Chengixang Liu ; Ruifeng Xu ; Lin Gui
Author_Institution :
Key Lab. of Network Oriented Intell. Comput., Harbin Inst. of Technol., Shenzhen, China
Abstract :
This paper presents a system to detect burst news on micro-blogging through mining burst words by incorporating features from micro-blog text, propagation periods and other characteristics. To avoid the event mixing problem, this system adopts originality estimation and event demising control to improve the burst news detection based on window-based incremental clustering. The experiments on Sina micro-blogging dataset achieve encouraging performance.
Keywords :
data mining; pattern clustering; social networking (online); text analysis; Sina microblogging dataset; burst event detection; burst news detection improvement; burst word mining; event demising control; event mixing problem; microblog text; originality estimation; propagation periods; window-based incremental clustering; Abstracts; Burst event detection; Event demise control; Micro-blogging; Originality estimation;
Conference_Titel :
Machine Learning and Cybernetics (ICMLC), 2013 International Conference on
Conference_Location :
Tianjin
DOI :
10.1109/ICMLC.2013.6890909
