copyright © 2006 tele atlas. all rights reserved. tele atlas and geographic names rob van essen –...
TRANSCRIPT
Copyright © 2006 Tele Atlas. All rights reserved.
Tele Atlas and Geographic Names
Rob Van Essen – EuroGeoNames Workshop, Utrecht, January 15th, 2007
Copyright © 2006 Tele Atlas. All rights reserved. 2
Overview
Tele Atlas Tele Atlas Map Database Data Sourcing Geographic Names Update process Quality Control Future
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 3
Tele Atlas : Background
Global geocontent provider serving wide range of markets
2005 revenue – €200 million/$238 million 2,300+ full time staff
and contract cartographers Shipped 6+ million
maps in 2005
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 4
Tele Atlas Partners
PNAVInternetMapping
ConsumerWireless
AutomotiveEnterprise/
Public Sector
Copyright © 2006 Tele Atlas. All rights reserved. 5
TA Global
Tele Atlas : Worldwide organization, Local Name handling
Tele Atlas
EMEAEurope
Middle-East
Africa
APACAsia-pacific
TANAThe Americas
DBO-N Scandinavia Great Brittain Ireland
DBO-CEE
DBO-DACH
DBO-W
DBO-SE
DBO-SW
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 6
Tele Atlas Database: Global Coverage
Tele Atlas
Tele Atlas Database
Data Sourcing
GeographicNames
Updating
Quality
FutureKEY Planned 2005 Planned 2006 - 08Current
Copyright © 2006 Tele Atlas. All rights reserved. 7
Tele Atlas Map Database: Data Model
Based on GDF: ISO/IS14825-2004 Real world objects: Features
Simple Features: Point, Line and Area Complex Features
Characteristics of real world objects: Attributes Simple and Composite attributes
Relations between real world objects: Relationships
Further development: ISO/TC204 WG3
Tele Atlas
Tele Atlas Database
Data Sourcing
GeographicNames
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 8
DBO-N Scandinavia Great Brittain Ireland
DBO-N Scandinavia Great Brittain Ireland
Data sourcing : Sources
DBO-N Scandinavia Great Brittain Ireland
Official institutions
Analogue dataMapsAtlasesCadastre maps
Field Update reports Mobile mapping …
Digital dataShape fileSatellite imageLists (xls, csv, txt, …)
Third parties Field surveyTele Atlas
Tele Atlas Database
Data Sourcing
GeographicNames
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 9
Geographic Names : Which Features?
Roads (streetnames, route numbers, …)
Administrative areasBuilt-up areasWater areasLand Cover, Land Use areas (parcs, islands, industrial areas, rest area grounds, Forest, …)
POI’s (Points Of Interest)
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 10
GeoNames: Some Statistics. Street names in EU database per language
Basque: 11.900 (ESP) Catalan: 47.000 (AND + ESP) Dutch: 182.000 (BEL + NLD) English: 338.000 (UK and part of IRL) French: 1.135.000 (BEL, CHE, FRA, ITA + LUX) Galician: 8.700 (ESP) German: 536.800 (AUT, BEL, CHE, DEU + ITA) Italian: 378.500 (CHE, ITA & SMR) Portuguese: 42.700 (PRT) Spanish: 201.000 (ESP) Valencian: 23.000 (ESP) Welsh: 2.900 (UK)
Total: 2.907.500 Street names Change Rate: 5%-15% yearly
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 11
Names: A Composite Attribute
NameName languageName typeName subtypeName components Character setTranscriptionsMeta data (Name origin, Rejection code, CreateDate, EditDate)
Tele Atlas
TA Map DB
Data Sourcing
GeographicNames
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 12
Names : Name Types and SubTypes
Name Types: Official Name Alternate Name Brandname Exonym Route Number …
Name Subtypes Highway name Tourist road national Tourist road regional Tourist road nature …
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 13
Geographic Names : Name Components
Body Key Prefix Suffix Pre-directional Post-directional Exit number Surname Article/preposition
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 14
Geographic Names : Character sets
Tele Atlas
Tele Atlas Database
Data Sourcing
GeographicNames
Updating
Quality
Future
Internal use of Windows character setsSpecific character set per set of countries
Benelux: Windows code page 1252 (Latin I) Poland : Windows code page 1250 (Central
Europe) Greece : Windows code page 1253 (Greek) Estonia : Windows code page 1257 (Baltic) … Future: UniCode
Products: UniCode
Copyright © 2006 Tele Atlas. All rights reserved. 15
Geographic Names : Transcriptions
Tele Atlas
Tele Atlas Map Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Tele Atlas databases contain Name Transcriptions to support Text-to-Speech and Speech-to-Text conversions of Geographic Names
11 Lexicographers Transcription = symbolic representation of the
pronunciation of a name consisting of phonemes, which are symbols, representing a sound
Supported Phoneme Alphabets: LH+broad, IPA, Sampa UCL, StarRec-SAMPA and Nuance
Example: Gaston Crommenlaan /GAs.'2tOnK_'krO.m$n.lan/gAs.'2tO%~_'krO.m$.lan/
/GAs.'2tOnK_'krO.m$.lan/ (LH+broad) /GAs$%tON#"krO$m@n$la:n/gAs$%to~#"krO$m@$la:n/
GAs$%tON#"krO$m@$la:n/ (Sampa UCL)
Copyright © 2006 Tele Atlas. All rights reserved. 16
Update process
Update sources Individual changes
Customer Feedback Update Reports from field survey Authorities
Mass updates (e.g. new streetfiles from third party) Source Material double checked in local offices Automatic import is done Quality checking is done afterwards (on
content, amounts, …)
Challenge: Handling m->n Feature – Name relationship: Name Storage per type
‘Schelde’ as river name ‘Schelde’ as street name ‘Schelde’ as name of a restaurant
Tele Atlas
Tele Atlas Database
Data Sourcing
Attributes
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 17
Quality Control
Unique name DB :all new names are checked for correctness by phonetic department as part of the creation process of transcriptions
QC includes more than 500 rules Illegal character set Syntax on name components Features that should have a name …
Regression checking via statistics per release Example:The Prefix Part of an Alternative Name or Standard Name generally starts
with an alphabetic character, a numeral or a single quote (').A comma can never occur in the prefix part of a Name. The Prefix Part shall not start with a space and shall not contain '' (double quotes), '//', '/', 'PX:', 'SN:', 'ON:' or any of these characters : [ ] = $ # % § < > ! ? \ ² ³ + : ; * µ £ & | @ " ( ) { } _ ¤ ^ (circumflex accent �without alphabetic character) ¨(diaeresis without alphabetic character) ´ (acute accent without alphabetic character) ` (grave accent without alphabetic character) ~ (tilde without alphabetic character).
Tele Atlas
Tele Atlas Database
Data Sourcing
Geographic Names
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved. 18
Future
Tele Atlas welcomes European initiatives for the creation of official uniform geographic name repositories made available to the European public and industry
Inventory of names (flat list) Categorized per language, per object,
per type, … Incl. geographic reference (e,g,
coordinate, Agora-C Location Reference (ISO/CD17572-2006, map)
Incl. change indicator (new, deleted, old, time)
And is willing to contribute to its design
Tele Atlas
Tele Atlas Databse
Data Sourcing
Attributes
Updating
Quality
Future
Copyright © 2006 Tele Atlas. All rights reserved.
Thank you for your attention!
Rob van EssenDirector Strategic Research, Tele Atlas [email protected]