• Aucun résultat trouvé

A Semantic Retrieval System for Remote Sensing Web Platforms

N/A
N/A
Protected

Academic year: 2021

Partager "A Semantic Retrieval System for Remote Sensing Web Platforms"

Copied!
16
0
0

Texte intégral

(1)

A Semantic Retrieval System for

Remote Sensing Web Platforms

(2)

Semantic

Semantic is the meaning and logic part

of the data

(3)

Semantic

Reducing the semantic gap between services providers and end-users thank to Ontologies

(4)

EOROntology

Semantic conceptual model for triple store.

Definition of what a processing chain is. What is it made of?

Nys, G.-A., Kasprzyk, J.-P., Hallot, P., & Billen, R. 2018. Towards an ontology for the structuring of remote sensing operations shared by different processing

chains. ISPRS - International Archives of the Photogrammetry, Remote

(5)

Natural Language Processing

Natural Language Processing

How to program computers to process and analyze large amounts of natural language data.

(6)

Workflow

(7)

Example

Query expressed in a natural language formulation :

How many trees are in a forest in Enschede ?

(8)

PoS Tagging

Part of Speech Tagging

process of marking up a word in a text (corpus)

as corresponding to a particular part of speech

How many trees are in a forest in Enschede ?

WRB NNS NN JJ VBP IN IN SYM NNP DT

(9)

Lemmatization

Lemmatization

process of grouping together the inflected forms of a word so they can be analysed as a single item

(10)

Query Expansion

3 possibilities:

1. Reference thesaurus 2. Spatial content

(11)

Reference thesaurus

Need to structure remote sensing words in a

shared common and well-described format. There is no remote sensing thesaurus.

(12)

Subgraph processing

Subgraph extraction and reconstruction

Simple Knowledge Organisation System

Edges definitions

Tree Plant

Pine Oak

(13)

Spatialisation

Geographic content

Contextualisation based on toponyms and place names: Nearest neighbours

(14)

Results

Weighted average based on the occurences

(15)

Prototype

(16)

geomatics.ulg.ac.be Coauthors FACULTY OF ARCHITECTURE LNA - DIVA PROF. PIERRE HALLOT FACULTY OFSCIENCES GEOMATICS RESEARCHASSOCIATE DR. JEAN-PAUL KASPRZYK FACULTY OFSCIENCES GEOMATICS RESEARCHASSOCIATE GILLES-ANTOINE NYS FACULTY OFSCIENCES GEOMATICS H U PROF. ROLAND BILLEN

Références

Documents relatifs

Association Between Query Terms and Ontology Input Concept or Relation in Ontology determinants health determinants increase ‘has positive effect on’ some.. risk ‘is disposition

When lecturers publish contents in an e-learning platform like a course degree or the bibliography for a given course, normally they do that either by uploading a document (MSWord,

As reading the API and pushing the information into the ontology does not depend on the API itself but on the input files, an API developer needs only one populator per

In this work, we explore the task of linking sets of labels from an arbitrary resource to a structured knowledge base, a problem we consider at the crossroad between word

For example, in our semantics, the answer to the query “hall discovered every moon” is computed by checking to see if, for every

У наступному розділі розглянуто розширення системи ОДСП для проектування та синтезу програм мовою Pig Latin з метою аналізу великих наборів даних

In this study we combine industry & academic (grey) literature resources, including business case descriptions, with interviews held with founding team members

With large, full text databases now commonplace, both IR and bibliometrics have also benefitted from developments in natural language processing (NLP) and