Database Reference
In-Depth Information
The Analytical Process in a Mobility Data Mining Platform
To give an idea of how the analytical process we described above concretely
maps into a dedicated analytical framework, we show here how the M-Atlas
system allows us to handle such complexity by means of an SQL-like language
called DMQL. All the analysis presented in this chapter can be expressed in
such language. Due to space limits, we cannot show here the corresponding
queries for all examples. We show how the previous process can be described
through DMQL. A more detailed introduction to the system is presented in
Chapter
7
.
First, the OD matrix is extracted according to a spatial tessellation and then
the trajectories between a given origin and several destinations are retrieved
from the data set:
CREATE MODEL MilanODMatrix AS MINE ODMATRIX
FROM (SELECT t.id, t.trajectory FROM TrajectoryTable t),
(SELECT orig.id, orig.area FROM MunicipalityTable orig),
(SELECT dest.id, dest.area FROM MunicipalityTable dest)
CREATE RELATION CenterToNESuburbTrajectories USING ENTAIL
FROM (SELECT t.id, t.trajectory
FROM TrajectoryTable t, MilanODMatrix m
WHERE m.origin = Milan AND
m.destination IN (Monza, ..., Brugherio))
The selected trajectories are then clustered to extract groups of trips with
similar characteristics. In the following query the route similarity function is
used:
CREATE MODEL ClusteringTable AS MINE T-CLUSTERING
FROM (Select t.id, t.trajectory from CenterToNESuburbTrajs t)
SET T-CLUSTERING.FUNCTION = ROUTE_SIMILARITY AND
T-CLUSTERING.EPS = 400 AND
T-CLUSTERING.MIN_PTS = 5
The extraction of cluster specimens from a specific day of the week and the
classification of new trajectories are performed by the following queries:
CREATE MODEL WednesdaySpecimens AS MINE SPECIMENS
FROM (SELECT id, trajectory, cid FROM WedTrajsToClusters)
SET SPECIMENS.MAX_DISTANCE = 750 AND
SPECIMENS.METHOD = ROUTE_SIMILARITY
CREATE TRANSFORMATION ClassifiedTrajectories
USING SPECIMENS_CLASSIFIER
