lod2 plenary vienna 2012: wp9a - lod for a distributed marketplace for public sector contracts
Post on 01-Nov-2014
650 Views
Preview:
DESCRIPTION
TRANSCRIPT
EU-FP7 LOD2 WP10 – 22.-23.9.2011. 02.09.2010 . Page 1 http://lod2.eu
Creating Knowledge out of Interlinked Data
http://lod2.eu
WP9a – LOD2 for a Distributed
Marketplace for Public Sector
Contracts
Plenary Meeting Vienna 21-23, March 2012 Vojtěch Svátek (UEP)
Collaborative Project 2010-2014
in Information and Communication Technologies
Project No. 257943
Start Date 01/09/2010
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 2 http://lod2.eu
Creating Knowledge out of Interlinked Data
1. Overall goals and status
2. Partners involved, tasks, deliverables and milestones
3. Achievements in M13-M18
4. Future plans
Agenda
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 3 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Explore and demonstrate the application of linked data principles for procuring
contracts in the public sector
• Provide best practices and (substantial) proof of concept for building the distributed
data platform
• Implement matchmaking and analysis services applicable on such a platform
• The use case (and WP) only started in M13, within the LOD2 Enlargement project
• Associated to WP9 in addressing government data• special focus
• association to (linked) commerce data
Overall goals and status
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 4 http://lod2.eu
Creating Knowledge out of Interlinked Data
• UEP 35 PMs
• I2G 5 PMs
• ULEI 5 PMs
• OKFN 3 PMs
Although most realization activities depend on UEP (University of Economics,
Prague), close collaboration with other partners is a must
• Support for use of individual technological components of the LOD2 Stack (currently:
Virtuoso, early experiments with OntoWiki and Silk)
• Public Procurement as one of integration use cases in WP6
• Participation in the linked data analytics – T9a.3, also related to WP10 (Linked Data
Mining Challenge)
Partners involved
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 5 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Task 9a.1: Creating linked data for public sector contracts• Started in M13, currently the main focus (data extraction and publishing)
• Task 9a.2: Matching the demand of public sector bodies with linked commerce data• Starts in M25
•Task 9a.3: Analytics of linked data for public sector contracts• Starts in M37
Tasks
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 6 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Deliverable 9a.1.1 Framework for creating linked data in the domain of public sector
contracts (originally due M16)• Scope of the deliverable was significantly extended, which caused a delay
• Not only general framework and ontology+cookbook, but also data infrastructure implementation
and data processing
• Draft submitted to internal review in (early) M19
• Deliverable 9a.1.2 Web application for filing public contracts (M24)• Presently starting the design of specifications – will be one of main topics of the WP break-out
session
• The remaining 4 deliverables (due M36+) are related to matchmaking and analytical
services
Deliverables
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 7 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Public Contracts Ontology (PCO)
• Data Processing Framework
• Datasets Processed
• Supply to Linked Open Data Mining Challenge
• Case Study in Supplier-Side Modelling
Current Achievements
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 8 http://lod2.eu
Creating Knowledge out of Interlinked Data
Public Contracts Ontology
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 9 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Ontology design• Reuse of existing RDF and non-RDF schemas (TED, Good Relations, SKOS, …)
• Mappings (Call for Anything, LOTED)
• Modularity (EU, particular countries, …)
• Comprehensive ‘cookbook’ for LD designers, covering all important constructs of the
PCO
• Started discussions with people involved in similar projects• WESO Oviedo,
• LOTED
• Euroalert.net
• Possible future extensions• National modules
• Modelling detailed award criteria, restrictions for suppliers (important for match-making)
Public Contracts Ontology
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 10 http://lod2.eu
Creating Knowledge out of Interlinked Data
• An instance of Virtuoso was deployed and is being filled with data extracted from
Czech and British PC resources
• Currently being extended with focused extractors, cleaners, linkers (Silk), quality
assessment components, data aggregation and visualization
Data Processing Framework
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 11 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Public contracts, Business entities
• Use:• Matchmaking
• Data mining and analytical services
• We have:• Snapshot of Czech national data (governmental portal, local portals – Prague, Universities etc.)
Cca 60K contracts
• British public contracts data (ContractsFinder)
Cca 7K contracts
• We need• More data from other EU countries and specific institutions
TED (not all contracts, more in national portals, involvement of other partners desirable)
• Data on companies from national business registers
opencorporates.com
• How can you help?• A little - Describe public contracts datasets in your country into CKAN – e.g. the Data Hub
• A lot - Screen-scraping or structured extraction to RDF data according to PCO
Datasets Processed
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 12 http://lod2.eu
Creating Knowledge out of Interlinked Data
First, exploratory run of the Challenge
• Spring 2012: data gathering and preparation; workshop submission to a conference• Public contracts data linkable to LOD and other LD resources
• Late Spring 2012: data analyzed by participants
• Autumn 2012: challenge workshop taking place
Data Supply for Linked Data Mining Challenge (part of WP10)
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 13 http://lod2.eu
Creating Knowledge out of Interlinked Data
• As proof of concept of supplier-side modelling, a vertical ontology for the Renewable
Energy Products domain was designed• Collaborative design relying on a Protégé – OntoWiki pipeline
• An initial experiment in matching PC data with potential supplier data in this domain
was carried out (using Silk)
Case Study in Supplier-Side Modelling
EU-FP7 LOD2 WP9a – 21.-23.3.2012. Page 14 http://lod2.eu
Creating Knowledge out of Interlinked Data
• Upon approval of the proposed framework we will more extensively publish and
refine public contracts data using the data infrastructure• A web-based application for public contracts filing will be developed (presumably, as an extension of
OntoWiki)
• D9a.1.2
• Existing inventory of ontologies for describing the supplier side will be examined and
new additions proposed (following the example of Renewable Energy Products
Ontology)
• Longer-term plans will be discussed at the break-out session on Friday
• Especially what LOD2 Stack tools can be used and what datasets can be processed!
Future Plans (in T9a.1)
top related