An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise...
Transcript of An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise...
![Page 1: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/1.jpg)
An Linked Data Repository for Transportation Planning Data
Megan Katsumi, Mark Fox
Enterprise Integration Lab (eil.utoronto.ca)
University of Toronto Transportation Research Institute (uttri.ca)
University of Toronto
Friday, May 31, 2019
![Page 2: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/2.jpg)
Problem: Planning transportation infrastructure over a 30 year horizon• What will demand for public
transportation and roads be over the next 30 years?
• How do changes in transportation infrastructure affect travelers?
• What are the environmental impacts of growth?
2
Transportation Planning
![Page 3: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/3.jpg)
GTAModel V4.0Travel demand forecasting
Pop & Emp by Zone
Synthesize persons, households, cars & jobs
PORPOW PORPOS
TASHA• Activity generation• Activity scheduling
• Tour-based model choice• Auto allocation
• Ridesharing
Emme Road & Transit Assignments by Time Period
Converged?
STOP
No
Yes
Location choice for non-work/school activities
High-order transit P&R station choice
Surface transit speed updating
External Trips &Other Special Generators
![Page 4: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/4.jpg)
Challenges for transportation planning data
1. Data representation: • Multitude of transportation planning tools are in use by researchers and cities• No easy way to compare results as each has their own unique data modelsWe need a standard for transportation planning data!
2. Data storage:• Heterogeneity of data: Data required to support transportation planning is
available in different formats with different representations. Consequently, related data are often stored in isolation.• Wasted effort: Data that is cleaned and integrated for one task may not be
reused for othersWe need a more effective way of storing transportation planning data!
4
![Page 5: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/5.jpg)
Challenge #1: Data representation
• Problem: • Multitude of transportation planning tools are in use by researchers and cities• No easy way to compare or reuse results as each tool has its own unique data
models
5
![Page 6: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/6.jpg)
Solution: a standard for transportation planning data• The problem may be addressed with a standard that:• Facilitates interoperability between heterogeneous data
• Works with different tools, data formats• Is easily extensible: tools and approaches are always changing• Has a unique interpretation; incorrect and correct interpretations should be
clearly identifiable• Claim: an ontology can be used to specify a standard that will satisfy
these requirements
6
![Page 7: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/7.jpg)
Transportation Standards for Interoperability
• Data Standards• Transmodel• gtfs• ISO 19107 Road network model• ISO 14825 Geographic Data Files (GDF)
standard • Geographic Markup Language (GML)• DATEX II• W3C Vehicle Information Service
Specification • SENSORIS• ADASIS• …
7
![Page 8: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/8.jpg)
Existing data standards
• Don’t focus on transportation planning• Subject to ambiguity• Potential for misinterpretation• Correspondences with other standards unclear
• May support syntactic interoperability, but cannot support semantic interoperability
8
![Page 9: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/9.jpg)
What is Semantic Interoperability?
• The ability of computer systems to exchange data with unambiguous, shared meaning. • A requirement to enable machine computable logic, inferencing, knowledge
discovery, and data federation between information systems.• Is concerned not just with the packaging of data (syntax), but the
simultaneous transmission of the meaning with the data (semantics)
9
![Page 10: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/10.jpg)
Sources of Data Semantics
10
Documentation
Source Code
Ontology
Implicit
Explicit
![Page 11: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/11.jpg)
What is an Ontology?
• (More than) a reference model for the domain.• Answers the questions:• What are the core concepts and properties that span the city’s data?
• To what extent can we generalize them in a useful way?• What are the key distinctions?
• Can we formally define necessary and/or sufficient conditions (using properties) for something to be an example (member) of a concept?
• A precise, formal (logical language) representation that supports:• Reuse• Integration• Automated deduction
11
![Page 12: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/12.jpg)
How are Ontologies Used?
• Data Integration:• Ontology to serve as an interlingua• Data and systems may be mapped into the ontology to support exchange of
information• Automated Deduction• New information may be inferred based on the data and knowledge of the
domain formalized with the ontology.• Model Checking:• Data may be automatically validated against the ontology to check whether it
conforms to the definitions.
12
Application 1
Interlingua (Ontology)
Application 2
Application NApplication 3 . . .
![Page 13: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/13.jpg)
Ontology Components
Micro-Theory• Axioms/Rules• Deduction – answering questions
Definitions and Constraints• Class Definitions (in Logic)• Automated classification
Knowledge Graph• Classes and Properties• Taxonomy and Inheritance
• Households• Transportation Network• Vehicles
• Household is composed of at least one person who resides at the same address
• For each year above the age of 14, a member of a household will leave with a probability p(Age)
![Page 14: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/14.jpg)
Example: City Resident
• Toronto: “you are identified as a resident if you reside in, own property, or own or operate a business in Toronto” (311 Toronto). • Beijing: “all individuals holding the nationality of the People’s Republic of China
who [have] a domicile in Beijing and nowhere else. If the individual maintains a regular dwelling somewhere else, the more regular dwelling is considered their place of residence” (Li, 1991). • New York: “the place which an individual intends to be his permanent home –
the place to which he intends to return. It is the home with range of sentiment, feeling and permanent association. One must be domiciled in New York and maintain a home in New York, the time spent in the State is irrelevant” (McGladrey, 2009). • Germany: “a resident of Germany generally refers to an individual who has a
domicile in Germany or spends more than six consecutive months in Germany (habitual place of abode)” (Seidel, 2011).
14
![Page 15: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/15.jpg)
How to Express the Semantics of Resident?
“you are identified as a resident if you reside in, own property, or own or operate a business in Toronto”
TorontoResident is-a Residentand (residesIn.Toronto
or ownsPropertyIn.Torontoor operates.(Business
and hasAddress.(Address and inCity.Toronto)))
15
![Page 16: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/16.jpg)
Different Views of a Concept
16
TorontoResident
BeijingResident
MumbaiResident
• residesIn• ownsProperty• operates
• hasAddress• inCity
• Address• Business
Defined in terms of shared fundamental properties and concepts
![Page 17: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/17.jpg)
An ontology-based standard for transportation planning data• Requirements
üFacilitate interoperability between heterogeneous dataüMust work with different tools, data formatsüMust have a unique interpretation; incorrect and correct interpretations
should be clearly identifiableüMust be easily extensible: tools and approaches are always changing
17
![Page 18: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/18.jpg)
Ontologies for Transportation Data
18
ILUTE
ATIS
Transit
Parking
Complete Streets
iCity Ontologies TPSOISO WG11 NWIP
iCity Projects
![Page 19: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/19.jpg)
TPSO: Components
Units of Measure ActivityLocation SensorsTime Change
Household
ParkingLand UseBuilding
Person Organization Trip Contact
Transportation Network VehicleTransit Travel Cost Trip Cost
![Page 20: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/20.jpg)
Solution for Challenge #1
• Use ontologies to specify a standard for transportation planning data • Project 1.1 has been developing an ontology for the iCity project• This ontology is being applied to serve as the basis for a standard
for transportation planning (the TPSO) with ISO WG11
20
![Page 21: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/21.jpg)
Challenge #2: Transportation Data Storage
• Problem:• Heterogeneity of data: Data required to support transportation planning is
available in different formats with different representations. Consequently, related data are often stored in isolation.• Wasted effort: Data that is cleaned and integrated for one task may not be
reusable for others
21
![Page 22: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/22.jpg)
Solution: a more effective data repository
• The challenges for transportation data storage may be addressed with a repository that:• Interprets data semantics in a clearly defined, commonly agreed upon, and
unambiguous way• Captures and leverages the relationships between data• Tracks data provenance
• Claim: A linked data repository can satisfy these requirements
22
![Page 23: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/23.jpg)
Linked Data Repositories
Linked data repositories store datasets using semantic web technologies• Data is captured using a standard language (RDF)• A shared vocabulary, defined with an ontology, may be used to
provide semantics for the data• Things (vocabulary terms and instances) are identified uniquely, using
Uniform Resource Identifiers (URIs)• This creates the ability to link within and between datasets• Linked data can be queried using the terms defined in the ontology
23
![Page 24: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/24.jpg)
Linked Data Repositories
24
Alice
Alice
Bob
typeTorontoResident
H1
hasMember
TorontoResidenttype
performsTrip
T1Bob
Acme
employsemploys
![Page 25: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/25.jpg)
Traditional vs Linked Data RepositoriesFeature Traditional Repository Linked Data RepositorySupport different formats and schema
With preprocessing With preprocessing; some direct translation possible
Clear interpretation of data semantics
✘ ✓
Easily extensible ✘ ✓Capture the relationships between data
Sometimes* ✓
Track provenance Sometimes ✓Reason about provenance ✘ ✓Easy access to data Sometimes* ✓Data validation (QA) Sometimes* ✓Perform reasoning with the data ✘ ✓
25*If a central DB with a well-defined schema is used to store the data
![Page 26: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/26.jpg)
Solution for Challenge #2
• Use ontologies to represent the semantics of the data • The proposed TPSO standard may be leveraged to facilitate correct
integration and reuse of data.• Use a linked data repository and the ontologies to store
transportation planning data• Ongoing work (Year 5) • GUDR: the Global Urban Data Repository is designed to serve as a linked data
repository for all urban data
26
![Page 27: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/27.jpg)
GUDR: Global Urban Data Repository
“A linked data repository in which all urban data can be deposited, searched and retrieved.”
27
![Page 28: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/28.jpg)
GUDR Stack
Analytics Services
Semantic Services
Search Services
Depositor Services
Maintenance Services
Quintuples
Key Characteristics:1. Meta-Data2. Vocabularies/Ontologies
![Page 29: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/29.jpg)
Meta-Data• Each RDF triples
has meta-data attached to it.• The process of
depositing automatically attaches meta-data.
![Page 30: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/30.jpg)
Vocabularies/Ontologies• The repository
includes vocabularies/ontologies.• Depositors
choose which to use.
![Page 31: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/31.jpg)
Next Steps
• Standardization of TPSO with ISO/JTC 1 WG 11 Smart Cities an ongoing effort• Continue development and implementation of GUDR• Ongoing development and research, e.g. data validity
• Leverage the TPSO and GUDR to provide a linked data repository for transportation planning data
31
![Page 32: An Linked Data Repository for Transportation Planning Data · Megan Katsumi, Mark Fox Enterprise Integration Lab (eil.utoronto.ca) University of Toronto Transportation Research Institute](https://reader036.fdocuments.us/reader036/viewer/2022071217/604b197a2296946324416b6c/html5/thumbnails/32.jpg)
Questions?
32