CDS/ISIS (Computerized Information Service / Integrated Scientific Information System) is a generic information storage and retrieval system, specializing in the management of non-numerical databases, i.e. databases whose main content is text.

Developed by UNESCO until 2005, today CDS/ISIS has become a free technique used for the creation of agile databases with high performance data retrieval.

In the 1960s, with the advent of new technologies, mega-libraries and bibliographic agencies started projects to automate their services. As the programs used were independent for the various functions (input, sorting, search and output of data) the International Labour Organization (ILO) created the ISIS (Integrated Set of Information Systems) for internal use, with the purpose of integrating these programs.

In 1975, Giampaolo Del Bigio, a programmer at UNESCO, proposed that its Computerized Documentation System (CDS) be based on the ILO's ISIS. The new programme was named CDS/ISIS.

UNESCO went on to develop and disseminate the system. Distribution was only to relevant non-profit institutions.

In the late 1980s, UNESCO released the distribution of CDS/ISIS, free of charge, to all persons and entities, including for-profit ones.

In 1995, Del Bigio presented the first results of Winisis (CDS/ISIS Windows version), using the C++ language, at the CDS/ISIS International Congress in Bogotá, Colombia. Winisis became the great promise of the future. In 1998 the version 1.0 was officially launched and in November of the current year the official version Winisis 1.3 was presented.

In January 2001, version 1.4 was launched, arriving in Brazil shortly afterwards. Bireme1, makes version 1.5 available for download. Currently version 1.5 (Build 3) is available in December 2003.

The CDS/Isis standard is based on the ISO 2709 standard, which specifies the requirements for the bibliographic record exchange format, and has three essential concepts that differentiate it from other database managers

• variable-size records: especially useful for bibliographic bases due to the diversity and complexity of the document description;
• repetitive fields: used to treat data with an uncertain number of occurrences, such as authors and subjects;
• subfields: subdivision of the field content, which allows to give different treatment to each indicated part.

CDS/Isis also uses an index treatment system which makes searching extremely fast.

The system modules allow, mainly:

• Data Entry/Maintenance;
• Information retrieval in the forms required by the user;
• Generation of reports in the most diverse forms, merging with information retrieval or not;
• Data interchange (import/export) with/to other Documentation Centres;
• Momentary search operation for several users (when the system is installed in a network).

CDSISIS, as it is software used in several countries, can guarantee not only the efficiency and the amplitude of the system, but above all it can make possible a real exchange of data in which the information units can act independently in the services provided to their Users.