Table 1.
Keywords list.
Table 2.
Examples of two types of Weibo data in standard format.
Fig 1.
Example of traffic anomaly detection using taxi GPS data.
Blue rectangular area confines the region for detecting anomalous paths; colored paths (red, green and purple) illustrate anomalous paths.
Fig 2.
Framework for fusing traffic information from physical and social transportation data.
Table 3.
Three types of traffic events reported by Weibo.
Fig 3.
Illustration of method of generating searching regions.
Green paths are the target road referred to in the Weibo message; blue rectangles represent generated searching regions.
Fig 4.
Anomaly index of road segments within searching regions during time window when a social transportation message was posted (Case 1).
Table 4.
Traffic anomaly matrix of Case 1.
Fig 5.
Anomaly index of road segments within searching regions during time window when a social transportation message was posted (Case 2).
Table 5.
Traffic anomaly matrix of Case 2.
Fig 6.
Anomaly index of road segments within searching regions during time window when social transportation messages were posted (Case 3).
Table 6.
Traffic anomaly matrix of Case 3.
Fig 7.
Anomaly index of road segments within searching regions during time window when a social transportation message was posted (Case 4).
The blue triangle is the geocoded location of the landmark referred in the Weibo message.
Table 7.
Traffic anomaly matrix of Case 4.
Fig 8.
Temporal analysis of traffic events.
(a) The probability distribution of duration of anomalies. The probability distribution of time interval between Weibo post window and (b) the beginning of the anomaly period, (c) the end of the anomaly period and (d) the time window with most anomalous regions among Weibo messages posted during a traffic anomaly.