CORDIS
EU research results

CORDIS

English EN

Analysis System of Symbolic Official data

Project information

Grant agreement ID: IST-2000-25161

  • Start date

    1 January 2001

  • End date

    31 December 2003

Funded under:

FP5-IST

  • Overall budget:

    € 2 372 184

  • EU contribution

    € 1 900 009

Coordinated by:

FACULTES UNIVERSITAIRES NOTRE-DAME DE LA PAIX

Belgium

Objective

ASSO will design methods, methodology and software tools for the analysis of multidimensionnal complex data (numerical and non numerical) coming from very large databases in statistical offices and administrations. Symbolic analysis is an extension of statistical data analysis methods to more complex objects called Symbolic Objects. ASSO will improve the software developed in SODAS project in order to render it more operational and attractive following the users requests as well as to add new innovative methods and to meet the needs of statistical offices.

Objectives:
The general objective of ASSO is to design methods, methodology and software tools for the analysis of multidimensional complex data (numerical or non numerical) coming from very large databases in statistical offices and administrations. Symbolic analysis extends statistical data analysis methods to more complex objects. Individuals processed by these methods are complex in the sense they represent groups of individuals, featuring variation among each group of individuals. Within the context of the project, the complex objects are called symbolic objects. The aim of the ASSO project is to improve the software built previously in order to render it more operational and attractive following users requests, to add new innovative methods and to meet needs of statistical offices.

Work description:
ASSO project will allow adaptation of statistical concepts to the dynamical global environment, will suggest methodology of treatments and will give guidance on understanding the data. Special attention will be given to background knowledge and Metadata in order to provide a "meaning" to the data, thus avoiding false interpretation due to data misunderstanding. Special care will be given to user-friendly human/machine interface. New visualisation techniques will be added in order to make information more accessible and transparent. Users will be concerned from the beginning to the end of the project. They will define data type and problems. They will work with experts of the methods to suggest methodology. They will evaluate permanently the progress of the project and prototypes. They will evaluate the final product with benchmarks. The main features of the new project will be:
- to improve the building of symbolic data from a Relational Data Base or directly
- to help the user in analysing administrative data,
- to introduce metadata in the description of data,
- to add new visualisation facilities,
- to add new methods for the design or the analysis of symbolic data and help the user in analysis results, as neural networks or Bayesian classification tree.L%- to give tools for the study of quality, stability and robustness of the methods.

The project will be divided into nine workpackages: one concerning the management, one concerning exploitation and dissemination, two concerning the kernel of the software accessible by all the developers and four concerning methods and tools.

Milestones:
Month 4 Project Plan, Document Standard, Consortium Agreement
Month 5 User requirements and context of use, Scientific methods specification
Month 9 Methodology, Specification of modules
Month 12 Benchmark definition
Month 24 First version of the software
Month 29 Benchmark evaluation, Test reports
Month 36 Final Benchmark and scientific report, second version of the software.

Coordinator

FACULTES UNIVERSITAIRES NOTRE-DAME DE LA PAIX

Address

Rue De Bruxelles 61
5000 Namur

Belgium

Participants (13)

DECISIA (CISIA)

France

EUSTAT, EUSKAL ESTATISTIKA ERAKUNDEA / INSTITUTO VASCO DE ESTADISTICA

Spain

FACULDADE DE ECONOMIA DO PORTO

Portugal

INSTITUT NATIONAL DE RECHERCHE EN INFORMATIQUE ET EN AUTOMATIQUE

France

INSTITUT TES

Luxembourg

INSTITUTO NACIONAL DE ESTATISTICA

Portugal

NATIONAL AND KAPODISTRIAN UNIVERSITY OF ATHENS

Greece

RHEINISCH-WESTFAELISCHE TECHNISCHE HOCHSCHULE AACHEN

Germany

STATISTICS FINLAND

Finland

UNIVERSIDADE FEDERAL DE PERNAMBUCO

Brazil

UNIVERSITA DEGLI STUDI DI BARI

Italy

UNIVERSITA DEGLI STUDI DI NAPOLI FEDERICO II

Italy

UNIVERSITE PARIS-DAUPHINE

France

Project information

Grant agreement ID: IST-2000-25161

  • Start date

    1 January 2001

  • End date

    31 December 2003

Funded under:

FP5-IST

  • Overall budget:

    € 2 372 184

  • EU contribution

    € 1 900 009

Coordinated by:

FACULTES UNIVERSITAIRES NOTRE-DAME DE LA PAIX

Belgium