CESSDA Expert Seminar CESSDA Expert Seminar Odense, 11-12/9/2008 Presentation made by Dimitra Kondyli.

CESSDA Expert Seminar CESSDA Expert Seminar

Odense, 11-12/9/2008

Presentation made by Dimitra Kondyli

Metadata Handling for GSDB/EKKE

Data and metadata are disposed through the following systems:

1) EKKE/GSDB (Greek Social Data Bank)

2) Nesstar Server

3) SPSS

1) The main system for EKKE is the Greek Social Data Bank (GSDB) : we import data centrally into the main database.

2) The schema of the main database is comparable to DDI but has some additional fields. By using the Nesstar publisher we produce data and metadata in Xml files and then the file is imported into the main database. Additional fields can be added (Supplementary Documentation)

3) As EKKE is a research organisation and not exclusively an archive some datasets from the research production of EKKE are stored in SPSS until they are imported to the main database

All datasets in GSDB are in greek , most of the metadata are in english.

All datasets documented by using the Nesstar publisher are in english.(around 75% of our study description are in English).

Additional fields concern the following:

project’s management, subject matter databases etc.

This information do not correspond to the DDI schema.

For all import/export functions automatic tools are used to import data either from the database to the DDI or from the SPSS to the database.

How the data are captured

Nesstar – automatic metadata capturing from a datafile

GSDB – a. by using SPSS b. DDI XML

Desktop applications

More about these two different systems of archiving:  

GSDB and Nesstar server.

Capturing metadata in Nesstar – as we all know concerns the automatic metadata capturing from a datafile (the metadata captured through this way concern datafile description such as number of variables, number of cases e.t.c. and variable

description such as variable name, label, variable statistics, value domain).

Through the GSDB there two possible ways of capturing metadata:

There is an old software tool that captures metadata from SPSS datafile

A new software tool captures metadata directly from DDI XML ver 2.

Both these applications are desktop applications

Datafiles corrections and File versioning

We have file versioning for the maintenance of metadata. We usually provide the latest versioning

and keep to the archive old versioning of the datasets

Maintenance

Metadata storage

• Oracle database and Xml files

Metadata presentation

• Cessda Portal• Nesstart Server• Node for Secondary Processing (NSP) : supports two types of applications :

1. The desktop application has been built using the - Centura programming language. The database in an Oracle Database and

the operating system of the database server is Solaris

2. The web applications has been built using -Perl scripting programming language. The database is the same database

used in desktop application. - Web Server : Apache and the operating system of the web server is also

Solaris Both applications use Oracle database and Unix Solaris Operating system

An end point:

Most of the database transactions (inserts, updates, deletes) are done through the

desktop application. The web application is used mostly for dissemination reasons