Semantic-based Trajectory Data Mining Methods

1
Semantic-based Trajectory Data Mining Methods

• Vania Bogorny
• INE UFSC

2
A importância de considerar a semântica
SC
T3
T3
T2
T2
T1
T1
T4
T4
Padrão Geométrico
3
Geometric Patterns X Semantic Patterns (Bogorny
2008)

• There is very little or no semantics in most DM
  approaches for trajectories

Consequence

• Patterns are purely geometrical
• Difficult to interpret from the users point of
  view
• Do not discover semantic patterns,
• which can be independent of spatial location

4
Principal Problema Falta de semântica
Trajetórias Brutas (x,y,t)
5
Motivada por um Modelo Conceitualpara Trajetórias
6
Trajetória Metafórica (Spaccapietra 2008)
Time
position
institution
7
Modelagem Conceitual (EPFL, Suíça)

• Primeiro modelo conceitual para trajetórias
• STOP parte importante de uma trajetória do ponto
  de vista de uma aplicação, considerando as
  seguintes restrições
• durante um stop o objeto móvel é considerado
  parado
• O stop tem uma duração (tf - ti gt 0)
• MOVE parte da trajetória entre 2 stops
  consecutivos ou entre um stop e o início/fim da
  trajetória

7
8
The Model of Stops and Moves (Spaccapietra 2008)

• STOPS
• Important parts of trajectories
• Where the moving object has stayed for a minimal
  amount of time
• Stops are application dependent
• Tourism application
• Hotels, touristic places, airport,
• Traffic Management Application
• Traffic lights, roundabouts, big events
• MOVES
• Are the parts that are not stops

9
Modelo de Stops e Moves
0N list
Has
11
2N list
hasStops
11
From
11
01
11
To
01
0N
0N
10
Adicionando semântica às trajetórias usando STOPS
STOPS são dependentes da aplicação
1
2
Ibis Hotel 1000-1200
Museu Louvre 1300 1700
Torre Eifel 1730 1800
Aeroporto 0800 0830
3
Rótula 0840 0845
Congestionamento 0900 0915
Cruzamento 1215 1222
Aeroporto 0800 0830
11
Semantic Trajectories

• A semantic trajectory is a set of stops and
  moves
• Stops have a place, a start time and an end time
• Moves are characterized by two consecutive stops

12
Métodos para instanciar o modelo de stops e moves
e minerar trajetórias semanticas
13
Methods to Compute Stops and Moves

• IB-SMoT (INTERSECTION-based)
• Interesting for applications like tourism and
  urban planning

2) CB-SMoT (SPEED-based clustering) Interesting
for applications where the speed is
important, like traffic management 3) DB-SMOT
(DIRECTION-based clustering) Interesting in
application where the direction variation is
important like fishing activities
14
IB-SMoT (Alvares 2007a)

• A candidate stop C is a tuple (RC, ?C), where
• RC is the geometry of the candidate stop (spatial
  feature type)
• ?C is the minimal time duration
• E.g. Hotel - 3 hours
• An application A is a finite set
• A C1 (RC1 , ?C1 ), , CN (RCN , ?CN)
  of candidate stops with non-overlapping
  geometries RC1, ,RCN
• E.g. Hotel - 3 hours, Museum 1 hour

15
IB-SMoT
(Alvares 2007ª)

• Input candidate stops // Application
• trajectories // trajectory samples
• Output
• Method
• For each trajectory
• Check if it intersects a candidat stop for a
  minimal amount of time

Semantic rich trajectories
16
Schema of Stops and Moves
Tid Sid SFTname SFTid Sbegint Sendt
1 1 Hotel 1 0825
0840 1 2 TouristicPlace 3
0905 0930 1 3 TouristicPlace 3
1001 1420
Stops
Moves
Tid Mid S1id S2id geometry
timest 1 1 1 2
48.888880 2.246102 0841 1 1 1
2 48.885732 2.255031 0842
... ... ... ... ... ... 1 1 1 2
48.860021 2.336105 0904 1 2
2 3 48.860515 2.349018 0941
... ... ... ...
... 1 2 2
3 48.861112 2.334167 1000
Touristic Place
Hotel
Id Name Stars geometry 1 Ibis
2 48.890015 2.246100, ... 2 Meridien
5 48.880005 2.283889,
Id Name Type geometry 1
Notre Dame Church 48.853611 2.349167, 2
Eiffel Tower Monument 48.858330 2.294333,
3 Louvre Museum 48.862220
2.335556,
Alvares (ACM-GIS 2007)
17
Queries Trajectory Samples X Stops and Moves
Q2 How many trajectories go from a Hotel to at
least one Touristic Place?
SELECT distinct count(t.Tid) FROM trajectory
t, trajectory u, hotel h, touristicPlace p WHERE
intersects (t.geometry, h.geometry) AND
Intersects (u.geometry, p.geometry) AND
t.Tidu.Tid AND u.timestgtt.timest
Trajectory samples
Semantic Trajectories
SELECT distinct count(a.Tid) FROM stop a,
stop b WHERE a.SFTname'Hotel' AND
b.SFTname'Touristic Place' AND a.Tidb.Tid
AND a.Sid lt b.Sid
No Spatial Join
Alvares (ACM-GIS 2007)
18
Queries Trajectory Samples X Stops and Moves
Q1 Which are the places that moving object A has
passed during his trajectory?
SELECT Hotel as place FROM trajectory t,
hotel h WHERE t.id'A' AND
intersects (t.movingpoint.geometry,h.geometry) UN
ION SELECT TouristicPlace as place FROM
trajectory t, touristicPlace p WHERE t.id'A'
AND intersects (t.movingpoint.geomt
etry,p.geometry) UNION
SELECT SFTname as place FROM stop WHERE id'A
Alvares (ACM-GIS 2007)
19
Queries Trajectory Samples X Stops and Moves
Q2 How many trajectories go from a Hotel to at
least one Touristic Place?
SELECT distinct count(t.Tid) FROM trajectory
t, trajectory u, hotel h, touristicPlace p WHERE
intersects (t.geometry, h.geometry) AND
Intersects (u.geometry, p.geometry) AND
t.Tidu.Tid AND u.timestgtt.timest
Trajectory samples
Semantic Trajectories
SELECT distinct count(a.Tid) FROM stop a,
stop b WHERE a.SFTname'Hotel' AND
b.SFTname'Touristic Place' AND a.Tidb.Tid
AND a.Sid lt b.Sid
No Spatial Join
Alvares (ACM-GIS 2007)
20
Queries Trajectory Samples X Stops and Moves
Q4 Which are the Touristic Places that moving
objects have passed and stayed for more than
one hour? SELECT temp.name, count() AS
n_visits FROM ( SELECT t.Tid, p.name
FROM trajectory t,
touristicplace p WHERE
intersects (t.geometry,p.geometry)
GROUP BY t.Tid, p.name
HAVING count(t.)gt60) AS temp GROUP BY
temp.name
SELECT t.name, count(s.) AS n_visits
FROM stop s, touristicplace p WHERE
s.SFTidp.id AND (s.Sendt - s.Sbegint ) gt
60 GROUP BY t.name
No Spatial Join
Alvares (ACM-GIS 2007)
21
CB-SMoT Speed-based clustering
(Palma 2008)
Input Trajectory samples Speed
variation minTime Output stops and
moves Step 1 find clusters
Step 2 Add semantics to each cluster
22
Stops (Methods SMot and CB-SMoT)
23
DB-SMOT Direction-based Clustering (Manso 2010)

• Input trajectories // trajectory samples
• minDirVariation //
  minimal direction variation
• minTime // minimum time
• maxTolerance
• Output semantic rich trajectories
• Method
• For each trajectory
• Find clusters with direction variation
• higher than minDirVariation
• For a minimal amount of time

24
Resultados obtidos com os Métodos que Agregam
Semântica Trajetórias de Barcos de Pesca
25
Resultados obtidos com os Metodos que Agregam
Semântica Trajetórias de Barcos de Pesca
26
Works Summarized in this part of the Tutorial
Geometric Pattern Mining Methods (mining is on sample points) Semantic Pattern Mining Methods (Generate Semantic Trajectories using DM - mining is on Semantic Trajectories) Behaviour Pattern Mining and Interpretation Methods
Laube 2004, 2005 Hwang 2005 Gudmundson 2006, 2007 Giannotti 2007 Lee 2007 Cao 2006, 2007 Lee 2007, 2008a, 2008b Li 2010 Alvares 2007 Zhou 2007 Palma 2008 Bogorny 2009 Bogorny 2010 Manso 2010 Alvares 2010 Giannotti 2009 Baglioni 2009 Ong 2010
27
CONSTANT Modelo mais recente para Trajetórias
Semanticas (Bogorny et al. 2012)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)