Micro-blog event detection and tracking based on the incremental hierarchical density clustering of representative points in cloud computing environment

Micro-blog event detection and tracking based on the incremental hierarchical density clustering of representative points in cloud computing environment

Fung Han Nan Jadong

In order to extract news events from the large amount of real-time information generated by microblogging service platform, a set of micro-blog event detection and tracking algorithm in cloud computing environment is proposed. Firstly, a new weight calculation method based on micro-Bo forwarding number and comment number is used to represent the micro-Bowenben as a vector space model, and then the incremental hierarchical density clustering (rihdbscan) algorithm based on the representative point is adopted to extract the keywords and finally realize the detection and tracking of news events. The algorithm is deployed on the cloud computing platform Hadoop for a single node that cannot quickly and efficiently handle massive micro-blogging data. Based on the real data obtained on the Sina Weibo platform, the results show that the proposed method of weight value calculation is more than.

Micro-blog event detection and tracking based on the incremental hierarchical density clustering of representative points in cloud computing environment