Research of Food Safety Event Detection Based on Multiple Data Sources

摘要

Online event detection techniques are usually used in single data source. This paper analyzes event detection in the perspective of multiple data sources, combining news reports and microblogs. Detect events from news, combining microblogs to do event monitoring and early warning. Also improve feature selection methods for multiple data sources event detection. Finally, the methods are applied to the detection of food safety events and the results of the research show that event detection with multiple data sources is meaningful and valuable.