The use of Semantic Web Technologies in Spatial Decision Support Systems Adam Iwaniak Jaromar Łukowicz Iwona Kaczmarek Marek Strzelecki The INSPIRE Conference 2013, 23-27 June Wroclaw University of Environmental and Life Sciences Kon-dor
How to make GIS systems more intelligent? How to make their interface simpler and more intuitive in use?
Add mouse to CAD the 80s Line segment is represented by a pair of points P1 and P2 with coordinates (x1,y1) and (x2,y2)
Add RDB How to find buildings that have two floors? graphics in CAD system, attributes in relational databases (Access, Oracle,...), example: MGE, GeoGraphics, AutoCad Map. Features Name_of_feature parcels buidlings roads Buildings Function Nr_storeys id 16 2 1011 17 3 1012 18 1 1013
Add topology Spatial queries Select * from Counties where Counties.pop < 30000 and Counties are within distance 100 mi of Major Cities; Spatial operator
Spatial + RDB The development of technology that stores geometry and attributes in a single database allows to use the mechanisms of RDBMS in GIS. This resolve problems with: multi-user access data protection versioning transactions backup performance data integrity.
Use SOA architecture A new dimension of heterogeneity on a global scale A design for PCs with GUIs WFS Server WMS Server WMS Server Catalogue Server Geoportal
Maturity of SDI in 2013 Implementation of the ideas from the early 90s, developed in other realities, before Google era Different business model for distributing spatial data Inefficient method of implementing CS-W OGC catalog services OGC standards are not fully compatible with W3C Automatic WPS orchestration is broken 2 milion of GIS users in 2001.
Web 2.0 and VGI
3D
Linked Data Best practices for publishing and connecting structured data on the Web: Use URIs as names for things Use HTTP URIs, so that people can look up those names When someone looks up a URI, provide useful information, using the standards (RDF, SPARQL) Include links to other URIs, so that they can discover more things.
A new era of Internet access Android already has been activated on 900 million devices Hugo Barra, Google I/O 2013 Keynote Notes More than half of Internet users in Poland use the Internet via a smartphone, it's four times more than voice a year and ago. touch, 45% no of mouse, Internet users as a reason no for keyboard, buying indicate no big the screen use of maps and GPS, MEC Analytics & Insight People have easier access to smartphones than toilets, Zet Radio
Example 1 1. Find the nearest petrol station 2. Find the shortest route between point A and point B 3. Find international hotel in Florence Task does not seem to be difficult to implement when the questions are predefined and the data is stored in one database which model is well known.
Example 2 Find a building in Wroclaw that is owned by Adam Iwaniak There are three databases available: administrative boundaries, parcels and their owners and topographic database Classic GIS * 1. Find all parcels in Wroclaw, owned by Adam Iwaniak 2. Find buildings which are located in these parcels 3. Show a list of buildings and their addresses *required inference
Possible solutions Creation of specialized software e.g. for mobile apps for searching buildings in Polish cities Tag buildings with a name of the owner - Google The use of Semantic Web technologies Siri is an intelligent personal assistant and knowledge navigator which works as an application for Apple's ios. The application uses a natural language user interface to answer questions, make recommendations, and perform actions by delegating requests to a set of Web services /wikipedia/
Semantic Web Technologies and standards that allow machines to understand the semantics of the information on the Web: Ontologies and thesauri Reasoning based on ontologies RDF RDFa RDFS, OWL SPARQL SWRL
Metadata expressed in RDFa - annotating html An alternative method of distributing metadata
Publication of spatial planning documents using ontology-driven approach, I.Kaczmarek Wednesday, 9:40, room 5
Spatial Decision Support System Interactive, computer-based system designed to support a user or a group of users in achieving a higher effectiveness of decision making while solving a semistructured spatial decision problem (Abdullah A., 2005).
SDSS to obtain decision on development conditions Issuing the decision about building conditions is possible only in the case of fulfilling the following conditions: 1) at least one adjacent parcel, reachable from the same public road, is built on in a way that the requirements for the new buildings can be defined 2) the parcel has access to a public road 3) the existing or planned utilities are sufficient for the development goal 4) the parcel does not require obtaining a permit for changing the land use type from rural or forested to non-rural and nonforested
Architecture of SDSS An alternative method of inference and analysis
Tbox - domain ontology
SDSS - Abox PostGIS RDF Topology ABox RDF Spatial objects
Abox - instances
Application ontology
SDSS and GIS Knowledge base Reasoner PostGIS
Summary of SDSS Similar results can be achieved using GIS system Unfortunately, using GIS is much easier The problem of processing geometry and topology Inference in OWL and SWRL Possibility of integration of data from different sources Rules are described in a formal way - knowledge sharing and knowledge reuse No need to transfer all the information from the geospatial database, the area of analysis can be narrowed.
Intelligent Spatial Information Infrastructure Inteligent SII Thesauri Ontology Semantic Web Services CS-W+ Metadata for use SDI WFS WMS WPS CS-W Metadata for discovery GIS dgn GML SHP Oracle MS SQL PostGIS
Intelligent Spatial Information Infrastructure Spatial Data Infrastructure Retrieval, publication and access to spatial data WMS, clearing house, CS-W, data set Spatial Information Infrastructure Interpretation of data, analysis and processing Data harmonization, WFS, WPS, XML, GML, XSLT, object oriented Intelligent Spatial Information Infrastructure The ability to understand by humans and computers the data and services across systems significantly facilitating: linking, retrieval and integration of information automated orchestration of geospatial web services inference new facts based on existing ones Semantic interoperability, linked data, ontology, RDF, OWL, object, URI
Summary Research on Semantic Web in SDI has a great potential It is conducted by many leading academic centers, but the number of practical solutions is limited However there are attempts to extend existing SDI services with semantic layer, and use a new standard for geospatial data GeoSPARQL Semantic Web can be alternative, but today is in an initial phase of development
Summary 1. Change in GIS technologies is required, because: Huge number of naive users Change of the user interface and user experience voice and touch instead of mouse and keyboard Simple interface needs more intelligence behind 2. The popularity of smartphones could be one of the key drivers of research on Semantic Web technologies. 3. Making Open and Machine Readable the New Default for Government Information May 9th, 2013, President`s Obama Executive Order.
Thank you for your attention adam.iwaniak@gmail.com
Additional slides for discussion
102 hasnr parcel2 contains buidling2 2 hasaddress RDF Sienkiewicza 3 Jan Kowalski area1 hasname Wrocław cd:area1 cd:hasname Wrocław spatial:contains :parcel1 spatial:contains :parcel2 by I. Kaczmarek overlaps parcel1 contains contains building1 hasowner hasaddress Adam Iwaniak Iwaniak 101 Mokra134 cd:parcel1 spatial:contains :building1 cd:hasowner Adam Iwaniak cd:hasnr 101 spatial:overlaps :area1 cd:parcel2 spatial:contains :building2 cd:hasowner Jan Kowalski cd:hasnr 102 cd:building1 cd:hasaddress Mokra 134 cd:building2 cd:hasaddress Sienkiewicza
SPARQL query Select all the buildings in Wroclaw which belong to Adam Iwaniak PREFIX cd: http://rdf.parcelontology.com/ spatial: http://geovocab.org/spatial/ SELECT?building?address WHERE {?parcel cd:hasowner Adam Iwaniak.?parcel spatial:contains?building.?parcel spatial:overlaps?area.?area cd:hasname Wrocław.?building cd:hasaddress?address}. Answer: Building: budynek1, Address: Mokra 134
Building ontology in Protege by J.Łukowicz <Ontology1348457921722:isOwnedBy rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#JanKowalski"/> <rdf:type rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#Plot"/> <Ontology1348457921722:overlaps rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#building_03"/> </Ontology1348457921722:SpatialObject> </Ontology1348457921722:owns> <Ontology1348457921722:owns rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#plot_06"/> <rdf:type rdf:resource="http://www.w3.org/2002/07/owl#namedindividual"/> <rdf:type rdf:resource="http://www.w3.org/2002/07/owl#thing"/> <Ontology1348457921722:owns rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#plot_05"/> <Ontology1348457921722:owns rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#building_03"/> <rdf:type rdf:resource="http://www.semanticweb.org/ontologies/2012/8/ontology134845792172 2.owl#Owner"/> <Ontology1348457921722:owns> <owl:namedindividual rdf:about="http://www.semanticweb.org/ontologies/2012/8/ontology1348457921722.o wl#plot_15">
Rule in SWRL Building(?building), OwnerOfPlot(?owner), Plot(?plot), iscontainedby(?building,?plot), isownedby(?plot,?owner) -> owns(?owner,?building) SPARQL SELECT?budynek FROM <http://www.semanticweb.org/ontologies/2012/8/ontolog y1348457921722.owl#> WHERE {?owner rdf:about myont:adamiwaniak.?owner myont:owns?budynek.?budynek myont:islocatedin myont:wroclaw. }