An evolutionary event detection model using the Matrix Decomposition Oriented Dirichlet Process

作者：

Highlights：

• We propose a method to detect events from streaming textual data from social media.

• Introducing the event evolution phenomenon for event detecting.

• A novel evolutionary model is proposed to capture the dynamism behavior of events.

• Proposing a Bayesian nonparametric model to construct an event detection system.

摘要

•We propose a method to detect events from streaming textual data from social media.•Introducing the event evolution phenomenon for event detecting.•A novel evolutionary model is proposed to capture the dynamism behavior of events.•Proposing a Bayesian nonparametric model to construct an event detection system.

论文关键词：Event detection,Event evolution,Topic modeling,Social network analysis,Incremental clustering

论文评审过程：Received 20 June 2020, Revised 23 June 2021, Accepted 12 October 2021, Available online 21 October 2021, Version of Record 26 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116086