• Aucun résultat trouvé

Towards an Ontology for the Structuration of Remote Sensing Operations Shared by Different Processing Chains

N/A
N/A
Protected

Academic year: 2021

Partager "Towards an Ontology for the Structuration of Remote Sensing Operations Shared by Different Processing Chains"

Copied!
16
0
0

Texte intégral

(1)

ISPRS Symposium

Towards

structuring

Nys Gilles-Antoine Jean-Paul Kasprzyk Pierre Hallot Billen Roland ISPRS Symposium

Delft 2018

operations

structuring

Towards an ontology for the

structuring of remote sensing

structuring of remote sensing

operations shared by different

processing chains

(2)

Background

Increase of use of remote sensing

Increase of use of remote sensing

Increase of use of remote sensing

Increase of use of remote sensing

(3)

Background

New web platforms are created as market

place about remote sensing products

place about remote sensing products

Many trades find utility in remote

New web platforms are created as market

place about remote sensing products

place about remote sensing products

(4)

Research hypothesis

Creation of a formal ontology structuring the

proposition of services may improve the creation

proposition of services may improve the creation

of web platforms dedicated to remote sensing.

Implementing such an ontology in a NoSQL

graph database and providing

hypothesis

a formal ontology structuring the

proposition of services may improve the creation

proposition of services may improve the creation

of web platforms dedicated to remote sensing.

Implementing such an ontology in a NoSQL

graph database and providing applications

(5)

Hybrid ontology

Application ontology

dedicated to processing chains structuring

chains structuring

Domain ontology structuring the collections concepts

(6)

Hybrid ontology

Application ontology

dedicated to processing chains structuring

chains structuring

Domain ontology structuring the collections concepts

(7)
(8)
(9)
(10)
(11)

Architecture

Classic architecture (Jena Apache) Two different access point

(12)
(13)

Knowledge discovery

Studies on pattern recognition for rules discovery

1. Does Operation type A often follows Operation type B ?

2. How often?

3. Why ?

discovery

Studies on pattern recognition for rules discovery

(14)

Conclusion

Application Ontology

Processing chains -> elementary operations linking data Knowledge discovery and rules learning

Insufficient number of processing chains

> elementary operations linking data Knowledge discovery and rules learning

(15)

Future work

More metadata to describe the

Switching Datatype Properties

Switching Datatype Properties

dc:description : Semantic Web search engine for

services market place

describe the data

Datatype Properties to Object Properties

Datatype Properties to Object Properties

(16)

Gilles-Antoine Nys

ganys@uliege.be @GA_Nys

Feel free to contact us

Jean-Paul Kasprzyk jp.kasprzyk@uliege.be Pierre Hallot p.hallot@uliege.be @p_hallot Roland Billen rbillen@uliege.be @RolandBillen Paul Kasprzyk LNA/DIVA Labs

Références

Documents relatifs

By a binary tree we mean an unordered rooted tree in which every vertex has at most two

in collaboration with Miguel Couceiro and Jean-Luc Marichal University of Luxembourg... Graphical interpretation of the

Referring to the us- age of structured information on the Web, digital twin models of man- ufacturing data can be seen as knowledge graphs that constantly need to be aligned with

During the milling process, the module tries to keep the tool axis as close as possible to the surface normal and simultaneously tries to keep the milling machine joint positions

This methodological support includes two textbooks published by the authors, guidelines for the use of remote sensing data processing software in educational and research

Thus, the use of active teaching methods (search laboratory works, individual or group projects) contributes to the better assimilation by students of the methods and

– The ML Process class represents the whole procedure of data processing and it contains four main subclasses, namely, Data Preprocess, Feature Preprocess, Modeling, and Evaluation..

The integration process of the spatial technologies into the Semantic Web stack is undertaken by defining a new kinds of FILTERs for SPARQL queries and new kinds Built-ins for