Download presentation
Presentation is loading. Please wait.
1
6/2/2015Andrew Frank1 REVIGIS Review Meeting WP 3 Andrew U. Frank Geoinformation TU Vienna frank@geoinfo.tuwien.ac.at Overheads at: http://www.geoinfo.tuwien.ac.at/presentations/frank.htm
2
6/2/2015Andrew Frank2 Goal of WP 3 General focus of REVIGIS: combination, quality, revision WP3 focus: “Fusion and revision of strongly constrained spatial objects. To deal with uncertainty from multiple and untrusted sources in the definition of spatial objects. The objects are generally crisply recorded and of symbolical nature, …”
3
6/2/2015Andrew Frank3 Tasks in WP 3 3.1 revision in multiple source problem 3.2 fusion and revision issues in a multi-use, multi-quality context (Praxitec, U Laval) 3.3 General implementation issues (start in month 12) Changes from plan: partner NMT left the project, replaced by Praxitec, U Laval) The details of the cases to be studied were adapted to meet current commercial interest.
4
6/2/2015Andrew Frank4 TU Vienna: Work done at the Technical University Vienna related to the REVIGIS project: 1.Ontology of administrative systems 2.Revising administrative rule bases 3.Updating cartographic data with multiple sources 4.Location Based Systems 5.Separation of information for different uses (Gruenbacher)
5
6/2/2015Andrew Frank5 1.Ontology of administrative systems WP is focused on administrative data (WP 2 is more related to scientific data) Administration uses crisp, discrete data which is derived from scientific data For example: a parcel (partition of space, limited operations, precise boundary points) What is the ontology of administrative objects? Example: what is a parcel, a person, money
6
6/2/2015Andrew Frank6 Administrative ontology Ontology in scientific systems are direct h = f (a,b,c) Difficulties with object formation were discussed under segmentation in WP 2. Administration – as part of the socially constructed reality – has a more complex ontology: X counts as y in context z (John Searle) X are physical objects (or actions) Y are legal concepts (objects or actions)
7
6/2/2015Andrew Frank7 Revisions in administrative data Observations of x1, x2 – the physical objects which ‘count as’ Update of y1, y2 – the inferred legal concepts Consistency between the observations (documents received, actions recorded etc.) and the established administrative properties.
8
6/2/2015Andrew Frank8 Achieved Simulation of ontology (Ph.D. completed – Steffen Bittner Searle’s explanation was based on ‘collective intentionality’ We have replaced this by the (monopoly of) force by the state Framework in which revisions can be formulated. Formalization of constraints management and update/revision operator for database.
9
6/2/2015Andrew Frank9 Commercial connections Simulation of systems where technical and legal issues cooperate. Application under discussion: Road pricing with Ericson
10
6/2/2015Andrew Frank10 2. Updating and revising administrative rule bases Changes in the administrative rules: consistency of the rules Question: How to maintain data collections where consistency was tested under different sets of rules?
11
6/2/2015Andrew Frank11 Achieved Formalization of the cadastral law using in an executable, algebraic language (Haskell). Translation paragraph by paragraph (not reconstruction of the law) Law is formalized, Ph.D. text is in revision
12
6/2/2015Andrew Frank12 Open Questions Consistency of the initial rule set (text of law) How to maintain consistency during revisions (by the legislator) Formal framework established, needs connection with WP 1
13
6/2/2015Andrew Frank13 Commercial interest Intergraph was initially interested in the research as a method to build applications with strong legal bases quickly. Navratil is currently seconded to Government of Cyprus to advice on the construction of Geographic Information System and Cadastre.
14
6/2/2015Andrew Frank14 3. Multi-data sources Documentation of a case where measurements and available cartographic data must be merged.
15
6/2/2015Andrew Frank15 Achieved: Documentation of the case Commercial relations: Data is from a commercial application, result should be directly implementable
16
6/2/2015Andrew Frank16 U Laval Fusion and revision in a mulit-use, multi- quality context Stress on managing data quality descriptions (metadata) Separate presentation afterwards
17
6/2/2015Andrew Frank17 4. Location Based Systems Many position measured. Update discrete data Special problem: Determine on which road a vehicle drives.
18
6/2/2015Andrew Frank18 Achieved This question is in definition phase. Commercial interest: Road pricing - Ericson
19
6/2/2015Andrew Frank19 Conclusion The questions of revisions of databases are of high commercial interest. Issue: what is the relevant information Technical question: fusion of data, test for consistency violations. how to measure data quality? New question: separation of information in multiple Geographic Information Products.
20
6/2/2015Andrew Frank20 Plans for year 2: Formalization Evaluate the formalizations discussed in WP1 and see how it applies. We have used Reiter’s formalization before. Default reasoning can be translated easily into an algebraic system (which is the formalism we use). Practically, hierarchical decomposition and local computation is limiting the length of inference chains and limits the combinatorial explosion.
21
6/2/2015Andrew Frank21 Plans for year 2: Ontology Connect the ontology research TU Vienna did under the Chorochronos project to the questions of data quality and data fusion. Compare ontological issues coming up in the land use/land cover (WP2) with the ontology of cadastre.
22
6/2/2015Andrew Frank22 Problems Reaction in a EU project are slow in comparison with the change in focus of commercial operations. This is true for the interest of partners in a project but also for changes in the application areas. Adaptation of project targets – especially for application areas – must be done flexible.
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.