UPV/EHU-UPPA

Download Report

Transcript UPV/EHU-UPPA

Collaboration Proposal
Call for research projects 2012: UPV/EHU-UPPA
Content of this document
Brief description of the call for projects UPV/EHU –UPPA
2012
Rare Ocean Species Collector
Challenges
Brief description of the call for projects UPV/EHU –UPPA
2012
Call for projects UPV/EHU -UPPA
Aim of the call
Initiate and finance common projects in order to setup, in near future, bigger
and more ambitious (multidisciplinary) projects: European funding Framework
Program, POCTEFA, FEDER, …
Grant Period
1 year
Maximum Budget
6K euros
Main expenses
Travels, technical assistance, …
Deadline
10th October 2012
Possible partners
Marine Biology
•LRHA/IFREMER(France)
•AZTI (Basque Country)
Computer Science
•LIUPPA (France)
•UPV Teams (Basque
Country)
Rare Ocean Species Collector
Project Context
Collecting information (location, species, etc.) related to the
capture of rare* fishes and invertebrates in the ocean is done by
several institutes (IFREMER, AZTI, Museum of Natural History, La
Rochelle University, etc.) mainly thanks to the help of fishermen
Problems
No common methodology
No data sharing
No information indexing/retrieval
No data analysis (since input data are only few)
* Never/occasionally captured previously
Project Aim
Setup an easy-to-use solution able to
Standardize information representation
Reduce fishermen efforts
Increase and speed up data input
Provide collective knowledge and data analysis
Project Framework
Rare Ocean Species
Collector
Insert
(photo+Meta)
Collective
knowledge
Fish
fisherman
Approve + Analysis
Photo
Expert
Search
(Text+Photo)
MetaData
Annotate
(Text+Photo)
MMI Ontology
User
How it works?
Rare Ocean Species
Collector
Insert
(photo+Meta)
Collective
When observing a captured rare fish, a fisherman can
Approve + Analysis
knowledge
Fish
fisherman
take a snapshot on his boat using his cellular phone
Photo
and
upload it into the provided application
Expert
Search
(Text+Photo)
MetaData
Annotate
(Text+Photo)
MMI Ontology
User
How it works?
Rare Ocean Species
Collector
Insert
(photo+Meta)
Collective
knowledge
Fish
fisherman
Approve + Analysis
Photo
Expert
Search
(Text+Photo)
Annotate
(Text+Photo)
User
MetaData
A user can look for rare fishes using his/her
smartphone. The retreival is based on: photo,
keywords, or hybrid
MMI Ontology
How it works?
Rare Ocean Species
Collector
Insert
(photo+Meta)
Collective
A user can annotate
(send feedback, enrich
Approve + Analysis
knowledge
Fish
fisherman
description and photos, etc.) rare fishes.
Photo
Expert
Search
(Text+Photo)
Annotate
(Text+Photo)
User
MetaData
A user can look for rare fishes using his/her
smartphone. The retreival is based on: photo,
keywords, or hybrid
MMI Ontology
How it works?
Rare Ocean Species
Collector
An expert approves the inclusion into the
knowledge
base if inputCollective
data concern
Insert
(photo+Meta)
rare fishes.knowledge
Fish
fisherman
Approve + Analysis
Photo
Expert
Search
(Text+Photo)
MetaData
Annotate
(Text+Photo)
MMI Ontology
User
How it works?
Rare Ocean Species
Collector
An expert approves the inclusion into the
knowledge
base if inputCollective
data concern
Insert
(photo+Meta)
rare fishes.knowledge
Fish
fisherman
Approve + Analysis
Photo
Expert
Search
An
expert can use the collective
MetaData
(Text+Photo)
knowledge to anaylse the location of
specificAnnotate
rare fishes, trace its spatio(Text+Photo)
temporal migration, etc.
MMI Ontology
User
Challenges
Challenges
Acquisition
How to define and easy-to-use and quality input acquisition
protocol?
How to assist the expert in his/her decision when no
enough/relevant data is provided?
Challenges
Storage
How to define a standard representation able to consider
different semantics and meanings?
How to reconstruct geographical information (action, subject,
time, 3D space, anaphoric resolution) from existing non
structured documents?
Famille Centrolophidae – Centrolophe noir (Centrolophus niger Gmelin, 1788)1
action
capturer
Centrolophe noir
time
janvier 1999
Anaphore
specie
subject
Geographical
Information
3D
localization Tarnos par 15m de profondeur
1
Espèce déjà capturée en janvier 1999 devant Tarnos par 15m de fond
Challenges
Storage
How to enrich/link data with external sources (Web, PreviMer,
SHOM, etc.)?
Building multimedia oriented crawlers/wrappers?
Building ontologies from structured and non structured sources?
Measuring information quality
Challenges
Retrieval
How to provide multi-criteria query (photo and text)?
How to adapt user/expert/fisherman interface (vocabulary, screens,
services)?
How to identify good services and how to compose new ones?
Mashup based approach
Challenges
Applications
How to generate on-the-fly applications (for mining, for annotating,
etc.) ?
Mashup based approach
Promotion
How to attract people to use “Rare Ocean Species Collector”
application(s)?
Towards Game Theory?