An Ontology-Based Traffic Accident Risk Mapping Framework

1
An Ontology-Based Traffic Accident Risk Mapping
Framework

• Jing Wang and Xin Wang
• Intelligent Geospatial Data Mining Group
• Department of Geomatics Engineering
• University of Calgary, Canada
• Aug 24, 2011

2
Outline

• Overview of Road Accident Problem

• An Ontology-Based Traffic Accident Risk Mapping
  Framework

• System Implementation and Case studies

• Conclusions and Future Work

3
Road Safety Problem
1.18

• World Health Organization estimates 1.18 million
  people were killed by road accidents in 2002.
• (Source WHO, 2004)

• In Canada, about 3,000 are killed every year on
  roads.
• (Source www.RememberRoadCrashVictims.ca)

3,000

• In Alberta, the average time between collisions
  is 5 minutes.
• (Source Tay, 2006)

5
How to reduce traffic fatalities and serious
injuries on public roads?
4
Accident Concentration

• The occurrences of traffic accidents
• Seldom random in space and time, but form
  clusters in geographic space
• These accident concentration areas or locations
  have increased likelihood for an accident to
  occur based on spatial dependency of historical
  data.

5
Research Goal

• Generate road traffic accident risk maps showing
  the risk area
• Risk not only reflect the accident numbers but
  also the degree of danger

6
Weakness of Traditional Approaches

• 1. Handle accident analysis at data level
• Cannot generate different maps to meet users
  different needs
• e.g. a map for the downtown area or the rush
  hours only

• 4-year (1999-2002) accident statistics on the
  16th Avenue of Calgary with the same time
  interval of the day

7
Weakness of Traditional Approaches

• 2. Ignore the severity levels of accidents

Accident with Injury
Accident with property damage only
Fatalities and injuries put more strain on the
network.
8
How to Improve?

• How to help user select the proper dataset?
• Naive option
• - Translate users' requirements into traditional
  database queries SQL Select from
  TrafficAccidentTable where AccidentCondition
  Rain AND location DowntownArea..
• Second option
• - Handle users' requirements at the knowledge
  level
• Ontology (provides domain knowledge include the
  non-spatial and spatial concepts and definitions
  relevant to the traffic accident. )
• How to generate the traffic accident risk map?
• DBSCAN (Density-Based Spatial Clustering of
  Applications with Noise)

9
ONTO_TARM Framework
ONTOlogy-Based Traffic Accident Risk Mapping
Framework
10
Domain Ontology

• The Traffic Accident Domain Ontology (TADO)
• A formal description of the classes of concepts
  and the relationships among those concepts that
  describe traffic accidents.
• Based on a 7-tuple structure O D, C, R, A, HC,
  prop, att
• domain context identifier D
• Concept set C
• The relation identifiers R
• Attributes describe C and R A
• A concept hierarchy classification HC
• Function prop
• Function att

11
Traffic Accident Domain Ontology (TADO)
GeopoliticalRegion
GeographicalRegion
Thing
Accident_Records
RoadCondition
LightCondition
EnvironmentalCondition
RoadSurfaceCondition
WeatherCondition
12
Reasoner

• Input of the reasoner is a users goal, and the
  output is a set of properties dataset
• Decompose into one or more spatial and
  non-spatial tasks
• Assemble returned results
• Example
• Risk map for The accidents happened in rush
  hours with bad weather in downtown Calgary
• Subtask 1 (Spatial task) find the "downtown
  Calgary".
• Subtask 2 (Nonspatial task) find the rush hours
  with bad weather .
• Subtask 2.1 (temporal condition task) find the
  rush hours
• Subtask 2.2 (weather condition task) find the
  bad weather

13
findDowntownAreaTask

• Pseudocode of spatial query task
  findDowntownAreaTask

• Sub-task findDowntownAreaTask
• defgoal find Calgary Downtown Area
• Input
• (object (is-a City) (object?ci) (hasName
  "Calgary"))
• (object (is-a CitySection) (object?cs)
• (hasName "Downtown Area") (insideOf?ci))
• (object (is-a community) (object?co)
  (insideOf?ci)
• (belong-section?cs))
• Output
• (object (is-a ?community) (object? co))

14
findAccidentConditionTask

• Pseudocode of Nonspatial query task
  findAccidentConditionTask

• sub-task findAccidentConditionTask
• defgoal find Accident Conditions
• Input
• (object EnvironmentalCondition?ec
• (RoadSurface-condition "dry"), (RoadCondition
  "straight" "curve"), (WeatherCondition
  findSevereWeatherTask()) (LightCondition
  "artificial""nature"))
• (object TemporalCondition?tc
• (Interval? findRushHoursTask()))
• (object (is-a AccidentCondition) (object?ac)
  (include?ec tc))
• Output
• (object (is-a ?AccidentCondition) (object?ac))

15
Risk Index

• How to define the risks?

Assign different weights to accidents with
different severity levels With in a given
accident dataset D,

i - ith severity level n the total number of
different severity levels Count() - a function to
get the total number of accidents at that
level Wi - the weight assigned to the ith
severity level
16
Risk Index Model

• Converted into Equivalent Property Damage Only
  (EPDO) accidents
• EPDO W1 Fatal W2 Injury W3 PDO
• PDO property damage only crashes
• e.g. PIARC (Permanent International Association
  of Road Congresses) recommended formula
• W19.5 W23.5 W31 EPDO 9.5 Fatal
  3.5 Injury PDO

Different jurisdictions use different weighting
schemesModel Ratio Source
1 111 Simple Total Crash Count 2
9.53.51 PIARC 3 76.88.41 North Carolina
DOT 4 136.134.941 Ohio DOT
5 779.913.881 Transport Canada 6
1300901 Federal Highway Administration
DOT Department of Transportation
17
Modified DBSCAN for Traffic Accidents
Density-based Clustering for Traffic Accident
Risk (DBCTAR)
MinPts 5
Eps 40
MinRisk
10
RiskIndex W1 Count(Fatal) W2 Count(Injury)
W3 Count(PDO)
19
gt RiskIndex Threshold MinRisk
18
Main Interface
Global Setting
Menus
Quick Setting Panel
Tool bar
Map Area
Layer Control
Status bar
19
Case Studies

• Dataset
• Reported collisions on the road within Alberta
  province (770,000 records)
• Network street dataset is clipped from Street Map
  of North American in the ArcGIS 9.3 Media Kit
• Community boundary dataset is from census
  subdivisions
• Users goals
• Case 1 to find a risk map "at rush hours in the
  morning of downtown area of Calgary"
• Case 2 to find a risk map between 800-1000pm
  in the downtown area of Calgary
• Case 3 to find a risk map on the Deerfoot Trail
  in Calgary

20
Results
Risk model PIARC, MinRisk 8, Eps45, MinPts3

(Case 2) Risk map between 800-1000PM of Calgary
downtown area
(Case 1) Risk map in rush hours (730-900AM) of
Calgary downtown area
(Case 3)Risk Map of Deerfoot Trail (extract)
21
Results
Road Accident Risk Mapping Web Publishing Platform
22
Results

• MinRisk

A risk map generated based on users requirement
Calgary downtown area under snow condition,
published with online platform
23
Evaluation
Detail Comparison of two mapping results
Site 1
Site 2
Kernel Density Estimation (KDE) Result (radius
set to 40 meters and cell size is 10X10 meters.)
(A) KDE
Site1 site2
1999-2004 29 30
PDO 25 27
Injury 4 3
density estimation (per 100m2) 1.4 1.5
Risk index with DBCTAR 80.25 68.64
2005 5 2
Site 1
Site 2
(B) DBCTAR
23
24
Evaluation
99-04 99-04 99-04 99-04 99-04 2005 2005 2005
Fatality Injury PDO count Risk index Fatality Injury PDO
Site 1 2 10 44 56 1742.6 0 5 14
Site 2 1 31 78 110 1288.18 0 0 9
Site 1
Site 2
25
Conclusions

• Ontology is first time integrated into a traffic
  accident risk mapping framework to generate
  different risk maps based on users' goals
• A density-based spatial clustering method for
  traffic accident risk (DBCTAR) is proposed
• To demonstrate the framework, a prototype of
  proposed framework has been implemented
• The preliminary results from the case studies are
  promising

26
Future Work

• Improve ontology reasoner and map generator
• Provide recommendations for the weight model
• Adopt other properties in the risk index model
• Explanations from civil experts

27
References

• WHO, 2004. World Health Organization, World
  Report on road traffic injury prevention, World
  Health Day Publication.
• RememberRoadCrashVictims.ca 2009.
  RememberRoadCrashVictims.ca
• Anderson, Tessa K. 2009. Kernel density
  estimation and K-means clustering to profile road
  accident hotspots. Accident Analysis Prevention
  41, no. 3 (May) 359-364.
• Borruso, G. 2005. Network density estimation
  analysis of point patterns over a network,
  Osvaldo Gervasi, Marina L. Gavrilova, Vipin
  Kumar, Antonio Laganà, Heow Pueh Lee, Youngsong
  Mun, David Taniar, Chih Jeng Kenneth Tan (Eds.)
  Computational Science and Its Applications -
  ICCSA 2005, International Conference, Singapore,
  May 9-12, 2005, Proceedings, Part III. Lecture
  Notes in Computer Science 3482, 126-132.
• Flahaut, B., Mouchart, M., Martin, E.S., and
  Thomas, I. 2003. The local spatial
  autocorrelation and the kernel method for
  identifying black zones a comparative approach,
  Accident Analysis Prevention, 35, 991-1004.
• Okabe, A. and Yamada, I., 2001. The K-function
  method on a network and its computational
  implementation. Geographical Analysis 33 3, pp.
  271290
• Okabe, A. Satoh, T. and Sugihara, K. 2009. A
  kernel density estimation method for networks,
  its computational method and a GIS-based tool.
  International Journal of Geographical Information
  Science 23(1)7 - 32.
• Shino S. 2008, Analysis of a distribution of
  point events using the network-based quadrat
  method, Geographical Analysis 40 (2008), pp.
  3804000.
• Steenberghen, T., Dufays T., Thomas,I. and
  Flahaut. B. 2004. Intra-urban location and
  clustering of road accidents using GIS a Belgian
  example. International Journal of Geographical
  Information Science 18(2) 169 - 181.
• Xie Z., Yan, J. 2008. Kernel density estimation
  of traffic accidents in a network space..
  Computers, Environment and Urban Systems, 32, pp.
  396-406.
• Ester, M. Kriegel, H. Sander, J. Xu, X. 1996. A
  Density-Based Algorithm for Discovering Clusters
  in Large Spatial Databases with Noise. In Proc.
  Second International Conference on Knowledge
  Discovery and Data Mining 226-231. Portland
  AAAI Press.
• Wang, X., Gu, W., Ziébelin, D. and Hamilton, H.
  2010. An Ontology-Based Framework for Geospatial
  Clustering, International Journal of Geographical
  Information Science, 24(1) 1601 - 1630
• Gruber, T. R, 1993. A translation approach to
  portable ontologies. Knowledge Acquisition, 5(2)
  (1993) 199-220

28
Thank You !
28
29
System Implementation
Prototype Implementation
29
30
Ontology Implementation
Using Protégé OWL
31
Interface
32
Mapping
Find the suitable parameters
Limited to the Network
Publish
Evaluation
Buffer Intersect
Clustering
33
Spatial Clustering
MinPts 5
K-dist (p) distance from the kth nearest
neighbour to p
Eps 40
q
Sorting by k-dist (p)
34
Threshold - MinRisk
The kth-distance of p (k30)
35
Threshold - MinRisk
The risk index value of the Eps neighbours of p
(Eps100)
36

• rule (?X tadoinsideOf ?Y) (?Y tadoinsideOf
  ?Z) -gt (?X tadoinsideOf ?Z)