Title
Discovering geo-dependent stories by combining density-based clustering and thread-based aggregation techniques.
Abstract
•This work combines geo-clustering and content aggregation to detect unexpected behavior in a city.•Content analysis provide explanatory information to anomalous clusters.•Results show that the complementarity of content and geo-location is beneficial.•Feasibility to be deployed in real scenarios.
Year
DOI
Venue
2018
10.1016/j.eswa.2017.11.019
Expert Systems with Applications
Keywords
Field
DocType
Data mining,Crowd detection,Density-based clustering,Content aggregation,Event detection
Data science,Data mining,Metadata,Social media,Social network,Computer science,Thread (computing),Cluster analysis,Commodity hardware
Journal
Volume
ISSN
Citations 
95
0957-4174
2
PageRank 
References 
Authors
0.36
21
4