CERIF based research information platform for Flanders, Belgium

CERIF2006 is the data model of choice for this solution. CERIF is the european standard for research registration. For more information, please see www.eurocris.org.

Ingestion is done in CERIF2006 XML, the formal CERIF data exchange format. For the purpose of this solution, a local CERIF XML variant is used for handling a number of EWI-specific business rules. This CERIF XML definition work was done by Atira based on EWI's documentation of the rules. Also, a small application was delivered for supplementary data validation to overcome certain inherent shortcomings of XML schema.

At database level, data is multi-versioned CERIF2006; a project-specific invention devised in order to handle EWI's request for roll-back options and other features depending on history- and versioning ability. However, once data has been successfully added to the system, a pure CERIF2006 representation is used. This is the case for all public views, which are available in pure CERIF2006 only. Finally, the solution makes data available as an OAI data provider in CERIF XML.

More information about CERIF standard is available at www.eurocris.org.

Ingestion can be handled either by XML file upload or by OAI-PHM based harvesting from the individual institutions' own databases. Either way, data must be represented as CERIF XML.

Each supplier can access a central administrator's application, by which the supplier can manage its own data; roll back of submissions are among the data management options. This application is available for all submitted data sets, no matter if they were uploaded or harvested. The solution's OAI harvesting mechanism executes based on schedules, which are administered in the admin web application.

Authenticated users can access a graphical admin user interface. Further, data providers (appointed employees of the Flemish research institutions) can carry out data validation, administer roll-back and perform similar operations. EWI staff can also access the application, but for different administrative tasks - the administration of data providers being one of them.

The term describes how data for public access is exposed in views to the database. In the first version of the solution, this view supports public data exhibition in two different ways: It feeds the public portal and it feeds an OAI data providing mechanism.

The portal is the central way of making information about Flemish research available to the public. A set of search and filtering features are available, as is number of ways for users to save their results. The retained relations between all content objects becomes a major advantage here, allowing the user to browse effortlessly from researcher to organisation, from organisation to project, and so on.

Data providing allows any external party to retrieve data represented as CERIF XML over the OAI protocol, and industry standard. Business rules and logic

Business rules and logic in general is handled in two separate cores, advanced and simple. Advanced core handles writing, multi-versioned data sets, roll-backs and similar demanding operations. Opposite to this, simple core handles reading only from the the public views - a much different task, which also agues this architecture.

The description above focuses on how research data is handles in the solution. However, other content and other processes are handled too - among these are user data, news, portal pages and logs.

The FRIS research portal will be realized as a Java application using a number of industry standard frameworks and components. Related to database, those include the Hibernate ORM framework, and the primary structural components are Java and Sun's Java System Application Server in the enterprise edition. Database and transaction configuration is done through the Spring framework, and the primary web interface components are JavaServer Faces, Facelets for view rendering, and the Tomahawk utility components.