Transcript apan Xi`an 2007 presentation

基于数据网格的数据共享与集成
Data-Grid based application of data
integration and sharing
Yu Xinwen, CAF
Yamakawa, A., Kiura, T. Ninomiya, S.,
NARC, NARO
Contents



Situation of Agro-Data-Grid
Rice data service
Integration system: RiceMet
Situation of Agro-Data-Grid

MetBroker，DemBroker，ChitsuBroker，
ResourceServer，SoilBroker





Addressing the issue of heterogeneity, both data and
software.
data sharing
Easy for data integration
However, these are all environment related data
We need crop data service for Agro-Data-Grid
Rice data service-an attempt for
Crop data service

Digital data sets are continuously produced





In spreadsheet format
Locally stored, isolated
Hard to share among researchers
Even harder for data merging and integration
Data service is the basis to address these
issues



Merging rice data needs a flexible data
model.
Merged data can be shared and integrated
via consistent access
Data service provides an access for
integration with environmental data
Implementation of Rice data
service

EJB3: Simpler and more flexible than older
version




POJO: reusable components
Annotation
Dependency injection: simpler JNDI lookup
Less configurations

Rice Data service

2 Entity Beans: RiceDataBean, LocationBean


Mapping rice data and location data
1 session bean: RiceDataSessionBean

Business logic to manipulate these data
RiceMet - Ajax+Google Maps


By using google maps, experimental
locations are visible on the map
help user to select datasets based on these
locations
ZK: Ajax toolkit with Java






Ajax-based Rich User Interfaces.
Java based programming, no Javascript
Rich components
Event-driven Model
Server centric processing
Full support to Google maps
Structure of RiceMet

Hybrid of MVC and SOA model

http://www.agmodel.org/RiceMet/
Google Map
Controller
Browser
Model
AMeDAS
AMeDAS
MetBroker
AMeDAS
View
Crop Data
Service
Client
Web application
Crop DB
Services
Integration between rice data and
weather data
RiceDataSer
vice
Rice data
table
Location
table

Integrated data are
important to crop
data analysis and
modeling
MetBroker
気象官署
AMeDas

Thank you for your attentions