Event Detection from Social Data Stream Based on Time-Frequency Analysis

摘要: Social data have been emerged as a special big resource of rich information, which is raw materials for diverse research to analyse complex relationship network users and huge amount daily exchanged packages on Network Services (SNS). The popularity current SNS in human life opens good challenge discover meaningful knowledge from senseless patterns. It an important task academic business fields understand user’s behaviour, hobbies viewpoints, but difficult issue especially large volume data. In this paper, we propose method extract real-world events Data Stream using approach time-frequency domain take advantage digital processing methods. Consequently, work expected significantly reduce the complexity social improve performance event detection resource.