ITS-EU Large-scale information sharing for the ITS domain

Links in the slides:
* Parking site data in Ghent: https://linked.open.gent/parking
* Federated query with parking lots: http://client.linkeddatafragments.org/#datasources=http%3A%2F%2Fdata.linkeddatafragments.org%2Flinkedgeodata;http%3A%2F%2Flinked.open.gent%2Fparking%2F&query=%0A%23%20Selects%20the%20name%2C%20the%20real-time%20available%20and%20the%20geolocation%20of%20a%20parking%20lot%20in%20Ghent%20with%20more%20than%20200%20spaces%20available%0A%0ASELECT%20%3Fname%20%20%3Ffree%20%3FwktString%20WHERE%20%7B%0A%20%20%3FParkingGent%20a%20%3Chttp%3A%2F%2Fvocab.datex.org%2Fterms%23UrbanParkingSite%3E.%0A%20%20%3FParkingGent%20owl%3AsameAs%20%3FparkingLGD%20.%0A%20%20%3FParkingGent%20rdfs%3Alabel%20%3Fname.%0A%20%20%3FparkingLGD%20%20%3Chttp%3A%2F%2Fgeovocab.org%2Fgeometry%23geometry%3E%20%5B%20%3Chttp%3A%2F%2Fwww.opengis.net%2Frdf%23asWKT%3E%20%3FwktString%5D%20.%0A%3FParkingGent%20%3Chttp%3A%2F%2Fvocab.datex.org%2Fterms%23parkingNumberOfVacantSpaces%3E%20%3Ffree%20.%0A%20FILTER%20(%20%3Ffree%20%3E%20200%20)%0A%7D

Transcript

1. PUBLIC Constraints for a large-scale ITS data sharing system Pieter Colpaert Peter Van der Perre, Dieter De Paepe, Thimo Thoeye, Ruben Verborgh, Erik Mannens
3. A service between 2 systems? Your system Third party system Agree on a protocol Will determine which questions can be answered Can ask fixed set of questions to your system
4. Sharing data on the Web Your data ? ? ? ? ? ? = preparing your data for maximum reuse
5. When do we need to maximize reuse? All parking spaces occupied Parking space available
6. Option 1: a service https://api.{myservice}/nearestParkingLots
7. Option 1: a service https://api.{myservice}/nearestParkingLots ?long={long} &lat={lat}
8. Option 1: a service https://api.{myservice}/nearestParkingLots ?long={long} &lat={lat} &closeToTramToCenter
9. Option 1: a service https://api.{myservice}/nearestParkingLots ?long={long} &lat={lat} &closeToTramToCenter &freeSpaceOnArrivalTime=17:00
10. Option 1: a service https://api.{myservice}/nearestParkingLots ?long={long} &lat={lat} &closeToTramToCenter &freeSpaceOnArrivalTime=17:00 You cannot answer all questions for everyone
11. Option 2: Idea of the Web – power to the clients https://{mydataserver}/parkings Cacheable documents, linked together. Smart clients download set of documents, regardless of server borders. Constraints 1. Stateless interaction 2. Cacheable documents 3. A uniform interface over HTTP ⇒ Following links to discover more information
12. Option 2: remains functional for real-time data time Now
13. New problem: semantic interoperability 1. Regarding identifiers Is there a documented global identifier that we can use or reuse? 2. Regarding domain model How do we document a “Parking Site” for machines? How do we know when 2 documents mean the same?
14. 1 ⇒ URIs for identifying parking sites Official identifiers for parking sites in Ghent Ghent Sint Pieters Square: https://stad.gent/id/parkings/P10 2 ⇒ Unofficial URIs for DATEX2 created http://vocab.datex2.org
15. Example: real-time and historic parking status In the city of Ghent: https://linked.open.gent/parking/ Federated Web querying on the client: http://bit.ly/2jUNnES ⇒ You can program your client to solve any question
16. A call for joining forces 1. How do we get DATEX2 to publish HTTP URIs for their terms? 2. Let’s start publishing data as documents instead of services to enable smart clients ⇒ only way in which we are really going to make intermodal route planning happen
17. PUBLIC