Скачать презентацию Semantic-based Trajectory Data Mining Methods Motivada por Скачать презентацию Semantic-based Trajectory Data Mining Methods Motivada por

504ffbfe4f50cdbb4b23904fb34cb590.ppt

  • Количество слайдов: 26

Semantic-based Trajectory Data Mining Methods Semantic-based Trajectory Data Mining Methods

Motivada por um Modelo Conceitual para Trajetórias Motivada por um Modelo Conceitual para Trajetórias

Principal Problema: Falta de semântica Trajetórias Brutas (x, y, t) Dados Geográficos Geografia + Principal Problema: Falta de semântica Trajetórias Brutas (x, y, t) Dados Geográficos Geografia + Trajetória Bruta = Trajetória Semântica 3

Trajetória Metafórica (Spaccapietra 2008) end Time (Professor, EPFL, 1988 -2010) (Professor, Dijon, 1983 -1988) Trajetória Metafórica (Spaccapietra 2008) end Time (Professor, EPFL, 1988 -2010) (Professor, Dijon, 1983 -1988) (Lecturer, Paris VI, 1972 -1983) (Assistant, Paris VI, 1966 -1972) position institution begin 4

Modelagem Conceitual (EPFL, Suíça) Primeiro modelo conceitual para trajetórias: q STOP: parte importante de Modelagem Conceitual (EPFL, Suíça) Primeiro modelo conceitual para trajetórias: q STOP: parte importante de uma trajetória do ponto de vista de uma aplicação, considerando as seguintes restrições: n n q durante um stop o objeto móvel é considerado parado O stop tem uma duração (tf - ti > 0) MOVE: parte da trajetória entre 2 stops consecutivos ou entre um stop e o início/fim da trajetória 5 5

The Model of Stops and Moves (Spaccapietra 2008) STOPS q q q Important parts The Model of Stops and Moves (Spaccapietra 2008) STOPS q q q Important parts of trajectories Where the moving object has stayed for a minimal amount of time Stops are application dependent n Tourism application q n Hotels, touristic places, airport, … Traffic Management Application q Traffic lights, roundabouts, big events… MOVES q Are the parts that are not stops

Modelo de Stops e Moves Traveler location 0: N list Has 1: 1 Trajectory Modelo de Stops e Moves Traveler location 0: N list Has 1: 1 Trajectory 2: N list has. Stops 1: 1 0: 1 From 1: 1 Move Stop 0: N 0: 1 To ƒ(T) 1: 1 Is. In 0: N Place 7

Adicionando semântica às trajetórias: usando STOPS são dependentes da aplicação 1 2 Aeroporto [08: Adicionando semântica às trajetórias: usando STOPS são dependentes da aplicação 1 2 Aeroporto [08: 00 – 08: 30] Ibis Hotel [10: 00 -12: 00]] Museu Louvre [13: 00 – 17: 00] Torre Eifel [17: 30 – 18: 00] 3 Aeroporto [08: 00 – 08: 30] Rótula [08: 40 – 08: 45] Congestionamento [09: 00 – 09: 15] Cruzamento [12: 15 – 12: 22] 8

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

Métodos para instanciar o modelos de stops e moves e minerar trajetórias semanticas 1 Métodos para instanciar o modelos de stops e moves e minerar trajetórias semanticas 1

Geometric Patterns X Semantic Patterns (Bogorny 2008) n There is very little or no Geometric Patterns X Semantic Patterns (Bogorny 2008) n There is very little or no semantics in most DM approaches for trajectories Consequence: • Patterns are purely geometrical • Difficult to interpret from the user’s point of view • Do not discover semantic patterns, which can be independent of spatial location

Methods to Compute Stops and Moves 1) IB-SMo. T (INTERSECTION-based) Interesting for applications like Methods to Compute Stops and Moves 1) IB-SMo. T (INTERSECTION-based) Interesting for applications like tourism and urban planning 2) CB-SMo. T (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

IB-SMo. T (Alvares 2007 a) A candidate stop C is a tuple (RC, C), IB-SMo. T (Alvares 2007 a) A candidate stop C is a tuple (RC, C), where q RC is the geometry of the candidate stop (spatial feature type) q C is the minimal time duration E. g. [Hotel - 3 hours] An application A is a finite set A = {C 1 = (RC 1 , C 1 ), …, CN = (RCN , CN)} of candidate stops with non-overlapping geometries RC 1, … , RCN E. g. [Hotel - 3 hours, Museum – 1 hour] of 90

IB-SMo. T n (Alvares 2007ª) Input: candidate stops // Application trajectories n Output: n IB-SMo. T n (Alvares 2007ª) Input: candidate stops // Application trajectories n Output: n // trajectory samples Semantic rich trajectories Method: q For each trajectory n Check if it intersects a candidat stop for a minimal amount of time Jurere 09 -12 Ibis. H. 13 -14 Floripa. S 16 -17 of 90

Schema of Stops and Moves Tid SFTname SFTid 1 1 Hotel 1 1 2 Schema of Stops and Moves Tid SFTname SFTid 1 1 Hotel 1 1 2 Touristic. Place 3 1 3 Touristic. Place 3 Sbegint 08: 25 09: 05 10: 01 Sendt 08: 40 09: 30 14: 20 Stops Tid Mid S 1 id S 2 id geometry 1 1 1 2 48. 888880 2. 246102 1 1 1 2 48. 885732 2. 255031. . . . 1 1 1 2 48. 860021 2. 336105 1 2 2 3 48. 860515 2. 349018. . . … 1 2 2 3 48. 861112 2. 334167 Hotel Id Name Stars geometry 1 Ibis 2 48. 890015 2. 246100, . . . 2 Meridien 5 48. 880005 2. 283889, … Alvares (ACM-GIS 2007) timest 08: 41 08: 42 Moves 09: 04 09: 41. . . 10: 00 Touristic Place Id Name 1 Notre Dame 2. 349167, … 2 Eiffel Tower 2. 294333, … 3 Louvre 2. 335556, … Type Church geometry 48. 853611 Monument 48. 858330 Museum 48. 862220 15 1

Queries: Trajectory Samples X Stops and Moves Q 2: How many trajectories go from Queries: Trajectory Samples X Stops and Moves Q 2: 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, touristic. Place p WHERE intersects (t. geometry, h. geometry) AND Trajectory samples Intersects (u. geometry, p. geometry) AND t. Tid=u. Tid AND u. timest>t. timest Semantic Trajectories SELECT distinct count(a. Tid) FROM stop a, stop b WHERE a. SFTname='Hotel' AND b. SFTname='Touristic Place' AND a. Tid=b. Tid No Spatial Join AND a. Sid < b. Sid 16 Alvares (ACM-GIS 2007) 1

Queries: Trajectory Samples X Stops and Moves Q 1: Which are the places that Queries: Trajectory Samples X Stops and Moves Q 1: 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) UNION SELECT ‘Touristic. Place’ as place FROM trajectory t, touristic. Place p WHERE t. id='A' AND intersects (t. movingpoint. geomtetry, p. geometry) UNION … SELECT SFTname as place FROM stop WHERE id='A‘ 17 Alvares (ACM-GIS 2007) 1

Queries: Trajectory Samples X Stops and Moves Q 2: How many trajectories go from Queries: Trajectory Samples X Stops and Moves Q 2: 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, touristic. Place p WHERE intersects (t. geometry, h. geometry) AND Trajectory samples Intersects (u. geometry, p. geometry) AND t. Tid=u. Tid AND u. timest>t. timest Semantic Trajectories No Spatial Join 18 Alvares (ACM-GIS 2007) SELECT distinct count(a. Tid) FROM stop a, stop b WHERE a. SFTname='Hotel' AND b. SFTname='Touristic Place' AND a. Tid=b. Tid AND a. Sid < b. Sid 1

Queries: Trajectory Samples X Stops and Moves Q 1: Which are the places that Queries: Trajectory Samples X Stops and Moves Q 1: 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) UNION SELECT ‘Touristic. Place’ as place FROM trajectory t, touristic. Place p WHERE t. id='A' AND intersects (t. movingpoint. geomtetry, p. geometry) UNION … SELECT SFTname as place FROM stop WHERE id='A‘ 19 Alvares (ACM-GIS 2007) 1

Queries: Trajectory Samples X Stops and Moves Q 4: Which are the Touristic Places Queries: Trajectory Samples X Stops and Moves Q 4: 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. *)>60) AS temp GROUP BY temp. name SELECT t. name, count(s. *) AS n_visits No Spatial Join FROM stop s, touristicplace p WHERE s. SFTid=p. id AND (s. Sendt - s. Sbegint ) > 60 GROUP BY t. name 20 Alvares (ACM-GIS 2007) 2

CB-SMo. T: Speed-based clustering (Palma 2008) Input: Trajectory samples Speed variation min. Time Output: CB-SMo. T: Speed-based clustering (Palma 2008) Input: Trajectory samples Speed variation min. Time Output: stops and moves Unknown stop Jurere 09 -12 Step 1: find clusters Step 2: Add semantics to each cluster 2. 1: If intersects during t stop Ibis. H. 13 -14 Floripa. S 16 -17 of 90 2. 2: If no intersection during t unknown stop Tutorial on Spatial and Spatio-Temporal Data Mining (ICDM 2010)

Stops (Methods SMot and CB-SMo. T) 22 2 Stops (Methods SMot and CB-SMo. T) 22 2

DB-SMOT : Direction-based Clustering (Manso 2010) n Input: trajectories min. Dir. Variation min. Time DB-SMOT : Direction-based Clustering (Manso 2010) n Input: trajectories min. Dir. Variation min. Time // trajectory samples // minimal direction variation // minimum time max. Tolerance n Output: semantic rich trajectories n Method: q For each trajectory q Find clusters with direction variation higher than min. Dir. Variation n For a minimal amount of time of 90

Resultados obtidos com os Métodos que Agregam Semântica – Trajetórias de Barcos de Pesca Resultados obtidos com os Métodos que Agregam Semântica – Trajetórias de Barcos de Pesca 2

Resultados obtidos com os Metodos que Agregam Semântica – Trajetórias de Barcos de Pesca Resultados obtidos com os Metodos que Agregam Semântica – Trajetórias de Barcos de Pesca 2

Works Summarized in this part of the Tutorial Geometric Pattern Mining Methods (mining is Works Summarized in this part of the Tutorial Geometric Pattern Mining Methods (mining is on sample points) Laube 2004, 2005 Hwang 2005 Gudmundson 2006, 2007 Giannotti 2007 Lee 2007 Cao 2006, 2007 Lee 2007, 2008 a, 2008 b Li 2010 Semantic Pattern Mining Behaviour Pattern Methods (Generate Mining and Semantic Trajectories using Interpretation Methods DM - mining is on Semantic Trajectories) Alvares 2007 Zhou 2007 Palma 2008 Bogorny 2009 Bogorny 2010 Manso 2010 Alvares 2010 Giannotti 2009 Baglioni 2009 Ong 2010