Ligue des Bibliothèques Europeénnes de Recherche, Groupe des Cartothécaires de LIBER
TRANSLATE ENGLISH to Français, Deutsch, Italiano, Português, Español! Explanation
© LIBER and author
Published from: LIBER Quarterly, the journal of European research
libraries, ISSN 1435-5205, Vol. 8(1998), No 2. With permission from K.G. Saur Verlag, Munich, Germany
E-mail: Pierre van Nypelseer
Abstract
The traditional information diffusion channels strongly separate information production, information diffusion and information exploitation by the final users. New technologies bring those tasks together in a single shared distributed environ-ment. We report here on the RAINBOW project. RAINBOW is now completing a prototype giving access on the World Wide Web to statistical and administrative files with the ability to perform data extractions and combinations, to build thematic maps and to display them immediately. The prototype can be extended to support data production.
RAINBOW project objectives
The traditional information diffusion channels strongly separate information production, information diffusion and information exploitation by the final users. New technologies can provide a single set of tools to support all those tasks in a shared distributed environment. The only limit that the users will encounter is their access rights. As a result, information retrieval ceases to be a specific task and is integrated in the flow of information processing.
We report here on the approach taken by the RAINBOW project. RAINBOW is supported in part by the Telematics for Administrations sector of the IVth Frame-work Programme of the European Commission. The present paper however reflects the view of the project participants, not necessarily those of the Commission. The long term goal of RAINBOW is to build reusable application parts for administrati-ons, accessible over the Internet. As a preliminary study, RAINBOW is now completing a prototype giving access on the World Wide Web to statistical and administrative files with the ability to perform extractions, combinations, build thematic maps and display them immediately, as shown in this prototype.
Users
The current target users are the European Official Statistical Institutes, Local Administrations and the citizens. The Statistical Institutes need to share, harmonise and compare data all over Europe. The Local administrations want to access data about their territory distributed across many national, regional and local administra-tions. They must be able to extract and combine data on thematic maps to perform studies and support decisions in domains such as neighbourhood improvement, river basin management, environment protection. They are responsible for providing information to the citizen. They are also involved in administrative procedures that span several departments or several administrations. Those procedures could run very fast on the World Wide Web.
Functionalities
The user accesses the RAINBOW prototype through a standard WWW client. The first screen gives him the choice of databases. Databases are accessible in Belgium, Italy, England, Greece and Portugal. The user can switch between databases at any time and move information between them. He sees each database as pages in a document that is opened simultaneously on all the databases.
Within each database, the user can browse through the available maps by zooming and clicking. On each map he has the choice of what information to display: roads, bus lines, stores, movie theaters or thematic information such as the distribution of employment across the city.
He also has access to textual information, such as the opening hours of the stores, the program of the movie theaters, the departments of each administration or how to run an administrative procedure when necessary. The more original part of the project however is that he also has access to all the information that was used to derive the thematic maps. He can display the description of the survey from which the information was derived. In the survey, he can display the file structure, in the file structure, the definition and the codification of each field. In a codification he has access to the list of codes and their definition. He can navigate through that information, finding all the surveys that use a specific professional status classifica-tion and then all the thematic maps derived from those surveys. Draft screens are presented in Figures 2 to 5.
The user can also perform natural language queries, that will tell him which classifications, surveys or maps deal with "unemployment" for instance. The query is traditional, but the information retrieved has much more structure.
The user is not limited to a description of the sources of the information. He can also see how it was derived: from which fields of which survey, using which computation and which trancodification. Those derivation rules are accessible in the World Wide Web document as easily as the maps or the data descriptions.
Finally, the user can fill in his own derivation descriptions and ask the system to perform the operations immediately. The system is based on an interpreter that can execute the operations that the user has described. This is analoguous to a spreads-heet, where the user enters arithmetic operations that can be both displayed and executed. Database management systems use the same principle: the user enters the description of specific files that can then be filled with data and queried.
Benefits
For the application builder side, this opens the possibility of developing highly generic software. The descriptions are created when the survey is created, and can drive all the tasks from computer assisted interviewing to statistical analysis and diffusion. A single application can handle many surveys. Only descriptions are changed. Other applications can be built from the same parts. New parts can be added. The parts are tailored to specific needs by the descriptions. Figure 1 shows that the prototype is built from a Thesaurus part, a Data Structure Description part and a Data Aggregation part. A GIS part has been added later. The Data Structure Description part handles the description of maps as well as the description of statitical files.
For the data producer, this tool supports a very efficient documentation manage-ment strategy. The descriptions are used both by people as documentation and by computers to perform operations (data extractions, map building, table building). The documentation is a by-product of data production, and not a specific task at a specific cost, at least for the structure part. The descriptions are complete and accurate since they have been used to produce the data. The natural language descriptions can be reused. The same codifications and the same code descriptions are found in many data sets.
The user is empowered. Without help from computer specialists, statisticians and cartographers, he retrieve information on the World Wide Web, find a complete documentation for each data set, perform extractions and build himself the particu-lar table and maps he needs.
For information managers, information documentation and information retrieval are not specific tasks anymore. They merge with information processing and understan-ding. However, support remains necessary for complex documentation, complex retrieval, complex domain knowledge and multi-linguism
RAINBOW project results
Our long term goal is to build reusable application parts that are accessible over the Internet using the distributed object specifications of the Object Management Group.
RAINBOW is a preliminary investigation supported in part by the European Commission in the Telematics for Administrations sector. Its public results will be:
RAINBOW project consortium
The project is user driven. Most of the software producers are also software users. The project coordinator is the Ministere Wallon de l'Equipement et des Transports (Belgium). The project is managed by AITECH (Belgium), the evaluation manager is the Facultes Universitaires Notre Dame de la Paix (Belgium) and the quality auditor is CONISY (Belgium). The Official Statistical Institutes involved are the Statistisches Landesamt Berlin (Germany), the Instituto Nacional de Estatistica (Portugal), the Institut National de Statistique (Belgium) and the Swiss Federal Statistical Office. The administrations are represented by the Ministere Wallon de l'Equipement et des Transports (Belgium), the Ville de La Louviere (Belgium), Regione Lazio (Italy) and the National Center of Public Administration (Greece). The industry is represented by Bull (France), Nuova Telespazio (Italy) and AITECH (Belgium). The participating universities are the Facultes Universitaires Notre Dame de la Paix (Belgium), The University of Edinburgh (United Kingdom) and the National Technical University of Athens (Greece).
Back to GdC Homepage