Geospatial Big Data: Business Cases from proDataMarket
- 1. Geospatial Big Data Business Cases from proDataMarket Dumitru Roman dumitru.roman@sintef.no
- 2. Geospatial Big Data Property Data and the proDataMarket project Example Business Cases Data Marketplace
- 3. http://www.millennium-project.org/ Geospatial Big Data are societal opportunities
- 4. Geospatial Big Data Raster Vector Sensors Mobile It’s easier than ever to collect geospatial data, but how can we exploit these geospatial big data?
- 5. Example: Property data One of the most valuable datasets managed by governments worldwide Extensively used in various domains by private and public organizations
- 6. Challenges in working with property data • Difficult to access • Cross-sectors • Data is highly heterogeneous and possibly large • Data quality • Time-consuming integration • Lack of innovation • …
- 7. How can we innovate (and make money) with property-related (Open) Data?
- 8. proDataMarket project goals • To make property data more accessible, more usable and easier to understand • To make it easier for: • Property data providers to publish and distribute their data • Data consumers to find and access property data needed for their businesses 2.5 Years (2015-2017) €4.5M 20+ Datasets
- 9. proDataMarket deliveries 7 data-driven business products and services 1 data marketplace
- 10. Example business case #1 Objective evaluation of the real estate properties Business Intelligence companies (e.g. Cerved) Automation and cost-reduction in property valuations, new services Public administration Fact-driven social policy Real estate agencies Speed up evaluation of properties, more objective estimation of properties Property buyers/sellers Eliminate intermediaries
- 11. Example business case #1 (cont’) Objective evaluation of the real estate properties in Italy, by Istat Census Snapshot of Italy, socio- demographic data about: house (its characteristics), people of the family (personal data, education, profession, work / study place) people that live in house (guests) OpenStreetMap Point of interest of the city about transport, downtown, environment Cadastral report Property details (surface, cadastral category, quality status, age, ownership details) ~ 10M buildings The evaluation of real estate property An up-to-date, objective evaluation of the real estate properties in territories in Italy = Market price € ++
- 12. =+ SYNTETIC INDEX ISTAT = -0.23 - (0.12 * UNEMPLOYED) + (0.2 * HISTORIC_BUILDINGS) + (0.58 * GRADUATES_ON_RESIDENTS) + (0.6 * STUDENTS_ON_RESIDENTS) SYNTETIC INDEX POI = -0.5 + (0.15 * closest_metro_station) + (0.14 * closest_railway_station) + (0.24 * n_bus_stops_within_800m) + (0.6 * n_small_green_areas_pois_within_800m) + (0.02 * n_pedestrian_paths_within_1000m) - (0.05 * closest_airport) Example business case #1 (cont’) Objective evaluation of the real estate properties in Italy, by
- 13. Example business case #1 (cont’) Objective evaluation of the real estate properties in Italy, by Sample technical challenges Semantic data heterogeneity How to translate a point of interest into an OSM query? How to retrieve data from the whole Italy? Structural data heterogeneity How to compute indicators on different data structures? Messy data How to exclude from computation duplicated annotations of the same real-world entities?
- 14. • Stakeholders: • Public administration (e.g. FEGA in Spain) • Farmers and land owners • Intermediaries (e.g. service providers) • Problems: • Unfair grant assignment and expenditure on audits • Incorrect grant assignments • Features defined subjectively Example business case #2 Common Agriculture Policy (CAP) funds assignments in Spain, by
- 15. Cadaster Information Parcels and their features: surfaces, limits, slope…. EFAs & LEs Ecological focused areas and Landscape elements accurately defined using LIDAR Satellite Kind of crops, Health status, Set aside zones, Nitrogen fixing crops, CO2 fixing crops… Accurately defined CAP parameters objectively defined, Automated process to create new datasets related to CAP Funds, Less errors, Less audits and field visits… = CAP Funds ++ Fund assignment rules examples • Crop Diversification • Kind, density and surface of Ecological Focus Areas • Conditionality Example business case #2 (cont’) Common Agriculture Policy (CAP) funds assignments in Spain, by
- 16. 4) There are patterns: Groups, lines, isolated trees, etc. 5) Trees in line, hedges Non-aligned groups, copses 6) A viewer 2) Classified points by their height 1) Raw datasets, just points 3) Points are grouped: Yellow (soil), Green (trees), Orange (bushes) Example business case #2 (cont’) Common Agriculture Policy (CAP) funds assignments in Spain, by
- 17. Example business case #3 (cont’) Augment Reality (AR) for Property-related Data in Norway, by AR for buildings AR for underground infrastructure What’s the impact of a new building on its surroundings? Where are the underground pipes?
- 20. • A hard copy of 314 pages and as a PDF file • 6 Person-Months • Data collection with spreadsheets • Quality assurance through e-mails and phone correspondence Pains: Time consuming, Poor data quality, Static report without live updating • Live service • Efficient sharing of data • Simplified integration with external datasets • Live updating • Reliable access • … • Risk and vulnerability analysis, e.g. buildings affected by flooding • Analysis of leasing prices Report Reporting Service 3rd party services Example business case #4 Reporting state-owned real estate properties in Norway, by
- 21. https://datagraft.net 21 Linked Data Approach: DataGraft
- 22. DataGraft: Data Transformation and Knowledge Graph Publication Process • Interactive design of transformations • Repeatable transformations • Reuse/share transformations (user- based access) • Cloud-based deployment of transformations • Self-serviced process • Data and Transformation as-a-Service 22 Transform Generate RDF Ontology X Ontology X Ontology X Ontology mapping RDF Graph Raw Data Prepared Data Map Map Semantic graph database
- 23. Geospatial Data is BIGthing Innovation with property-related data in proDataMarket
- 25. 05.05.2016 25