using linked data in a heterogeneous sensor web: challenges, experiments and lessons learned

23
National Center for Supercomputing Applications University of Illinois at Urbana-Champaign Using Linked Data in a Heterogeneous Sensor Web: Challenges, Experiments and Lessons Learned Liang Yu and Yong Liu {liangyu, yongliu}@ncsa.illinois.edu

Upload: cybera-inc

Post on 24-May-2015

913 views

Category:

Education


2 download

DESCRIPTION

Presentation by Liang Yu during the Sensor Web Ontology and Semantics paper session of the Sensor Web Enablement workshop (held during the 2011 Cybera Summit).

TRANSCRIPT

Page 1: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

National Center for Supercomputing Applications University of Illinois at Urbana-Champaign

Using Linked Data in a Heterogeneous Sensor Web: Challenges, Experiments and Lessons Learned Liang Yu and Yong Liu {liangyu, yongliu}@ncsa.illinois.edu

Page 2: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Outline

Ø Motivation and Goal Ø Using Linked Data for Integration Ø Services for Consuming Linked Data Ø An Example: Visualization Ø Conclusion and Future Work

Imaginations unbound

Page 3: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Motivation

•  Interoperability as a grand challenge in a Virtual Environmental Observatory •  The need to integrating different sensor data, citizen sensing,

and other geospatial data etc. •  e.g., all river gage height values from all gages in Illinois

between 1st May and 3rd May 2011.

•  A heterogeneous Sensor Web is a reality •  Different protocols and formats, unknown semantics and no

links.

•  A paradigm shift: using Linked Data to build a Linked Sensor Web •  Linked data has the potential to provide a solution: URI naming

system, HTTP protocol, RDF, links to other resources. •  E.g. an OGC SOS usually has no links to other resources

Imaginations unbound

Page 4: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Objective: Data Integration in a Virtual Environmental Observatory

Imaginations unbound

WaterML

Page 5: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Outline

Ø Motivation and Goal Ø Using Linked Data for Integration Ø Services for Consuming Linked Data Ø An Example: Visualization Ø Conclusion and Future Work

Imaginations unbound

Page 6: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

A Simplified View of Linked Data

•  Obtain sensors and observation data from sensors in Illinois

Imaginations unbound

Illinois States Sensors in Illinois

Data Stream

http://sensorweb.ncsa.uiuc.edu/data/map/state/USGS/ Illinois

http://sensorweb.ncsa.uiuc.edu/api/sensordata/sites?within=http://sensorweb.ncsa.uiuc.edu/data/map/state/USGS/Illinois

http://sensorweb.ncsa.uiuc.edu/api/sensordata/observations?observedBy.within=http://sensorweb.ncsa.uiuc.edu/data/map/state/USGS/Illinois

Page 7: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Key Challenges

•  Publishing Linked Data •  Re-publishing existing plain data to semantically linked data. •  Linking potentially “linkable data” together and enabling complex

queries in a heterogeneous Sensor Web.

•  Consuming Linked Data •  Serving data in an OGC RESTful SOS-like service. •  Tracking the provenance of Linked Data to facilitate trust and

validation.

Imaginations unbound

Page 8: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Workflow of Publishing and Consuming Linked Data

Imaginations unbound

Ontology

Geometric Analysis

Linked Data

LDA

Tools

Raw Data Linked Data

Page 9: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Republishing Data to RDF (1)

•  Use XML as an intermediate format •  All non-XML data (Shapefile, Excel) are transformed to XML.

•  Use XSLT to convert intermediate XML to RDF/XML •  Three different annotations are implemented in XSLT

•  semantic annotation (rdf:type) •  outgoing links (to DBPedia), and •  provenance.

•  Both syntactic transformations and semantic annotations are performed

Imaginations unbound

Page 10: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Republishing Data to RDF (2)

Imaginations unbound

provenance

rdf:type

Outgoing link

Page 11: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Using Multiple Ontologies

Imaginations unbound

Page 12: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Creating Links by Geometric Analysis (1)

•  To discover and create links between entities based on their spatial attributes. •  Different from previous work

•  E.g.: http://www4.wiwiss.fu-berlin.de/bizer/silk/

•  Tools: Geotools (Java).

Imaginations unbound

Page 13: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Creating Links by Geometric Analysis (2)

Imaginations unbound

Page 14: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Provenance in Linked Data

•  Where did the data come from? How were they processed?

•  Using Open Provenance Model (OPM)

Imaginations unbound

<ssn:observedBy rdf:resource="http://sensorweb.ncsa.uiuc.edu/data/sensordata/sites/CUAHSI/NWIS/03339000"/> <opmo:wasDerivedFromStar rdf:resource="http://waterservices.usgs.gov/nwis/iv"/>

Sensor

Observation opmo:wasDerivedFromStar

opmo:wasDerivedFromStar

ssn:observedBy opmo:WasGeneratedBy

Page 15: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Outline

Ø Motivation and Goal Ø Using Linked Data for Integration Ø Services for Consuming Linked Data Ø An Example: Visualization Ø Conclusion and Future Work

Imaginations unbound

Page 16: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Developing SOS with Linked Data API (1)

•  An OGC RESTful SOS-like service over the integrated linked data •  Can be accessed by simple URLs. •  Is as flexible as SPARQL but with simpler syntax.

Imaginations unbound

q  Feature of Interest (http://sensorweb.ncsa.uiuc.edu/data/map/watershed/USGS/2009_0)

q  Single sensor site (http://sensorweb.ncsa.uiuc.edu/data/sensordata/sites/CUAHSI/EPA/MWRDSTOR:WW_39)

q  Collection of sensor sites (http://sensorweb.ncsa.uiuc.edu/api/sensordata/sites)

q  Single observation (http://sensorweb.ncsa.uiuc.edu/data/event/hail/noaa/2010/10908_2010-12-31T22:40:00)

q  Collection of observations (http://sensorweb.ncsa.uiuc.edu/api/sensordata/observations)

Page 17: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Developing SOS with Linked Data API (2) •  Obtain all the sensor sites within Illinois state.

•  http://sensorweb.ncsa.uiuc.edu/api/sensordata/sites?within=http://sensorweb.ncsa.uiuc.edu/data/map/state/USGS/Illinois

Imaginations unbound

items": [ {"_about": "http://sensorweb.ncsa.uiuc.edu/data/sensordata/sites/CUAHSI/EPA/MWRDSTOR:WW_39", "hasCode": "MWRDSTOR:WW 39", "hasLocation": {

"lat": 41.88185119628906, "long": -87.63558197021484, "type": "http://www.w3.org/2003/01/geo/wgs84_pos#Point"}, "hasName": "South Branch Chicago River @ Madison St.", "hasNetwork": "http://sensorweb.ncsa.uiuc.edu/data/sensordata/network/CUAHSI/EPA", "hasStream": "http://sensorweb.ncsa.uiuc.edu/api/sensordata/observations?observedBy=http://sensorweb.ncsa.uiuc.edu/data/sensordata/sites/CUAHSI/EPA/MWRDSTOR:WW_39", "type": "Sensor", "wasDerivedFromStar": "http://water.sdsc.edu/waterOneFlow/", "within": ["http://sensorweb.ncsa.uiuc.edu/data/map/county/USGS/Cook_County,_IL","http://sensorweb.ncsa.uiuc.edu/data/map/state/USGS/Illinois","http://sensorweb.ncsa.uiuc.edu/data/map/watershed/USGS/2009_2627"]},

Page 18: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Outline

Ø Motivation and Goal Ø Using Linked Data for Integration Ø Services for Consuming Linked Data Ø An Example: Visualization Ø Conclusion and Future Work

Imaginations unbound

Page 19: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

An Example Application (Visualization of SOS Results) •  Visualization Platform: WWT (World Wide Telescope)|Earth •  Sensor Data: river gage height values produced by gages in Illinois between

2011-05-01 and 2011-05-03. q  http://sensorweb.ncsa.uiuc.edu/api/sensordata/observations?observedBy.within=http://

sensorweb.ncsa.uiuc.edu/data/map/state/USGS/Illinois&_sort=observationResultTime.inXSDDateTime&min-observationResultTime.inXSDDateTime=2011-05-01T00:00:00-05:00&_page=0&max-observationResultTime.inXSDDateTime=2011-05-03T00:00:00-05:00&observedProperty=http://sensorweb.ncsa.uiuc.edu/data/property/USGS/NWIS:UnitValues/00065

Imaginations unbound

Page 20: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Outline

Ø Motivation and Goal Ø Using Linked Data for Integration Ø Services for Consuming Linked Data Ø An Example: Visualization Ø Conclusion and Future Work

Imaginations unbound

Page 21: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Conclusion

•  A “Linked Sensor Web” solution for integrating heterogeneous sensor data sources and geospatial data. •  A best practice of using W3C SSN ontology as well as other

domain ontologies. •  A method to discover and create links among spatial data. •  Using Linked Data API to provide OGC RESTful SOS-like

services •  http://sensorweb.ncsa.uiuc.edu/api-list.html

•  Tracking provenance data and managing them using OPM.

Imaginations unbound

Page 22: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Future Work

•  Towards Linked Geostreaming Data •  Tuning the performance •  Facilitating searching and crawling of Linked data •  End-to-end provenance management

Imaginations unbound

Page 23: Using linked data in a heterogeneous sensor web: Challenges, experiments and lessons learned

Imaginations unbound

Acknowledgements

Ø Microsoft Research Connections Ø  “Environmental Informatics” Program

Ø  Institute for Advanced Computing Applications and Technologies at the University of Illinois at Urbana-Champaign Ø  “Virtual Observatory for Sustainability of Intensively Managed

Environmental Systems” Project