Turning Roget’s Thesaurus into a Czech Thesaurus in a report on how a thesaurus of the Czech language was compiled on the basis of Roget’s Thesaurus, the following issues are covered:
1. Reasons for undertaking the thesaurus project-to redress the unbalance between the semasiological and onomasiological description of Czech by compiling a counterpart to the two large alphabetical dictionaries of Czech;
2. Strategy and philosophy, and the choice of the source text-combination of translation and original compilation; decision to use an available and well-proven model, a shorter version of Roget’s Thesaurus, to resolve the issue o f a classificatory system and format;
3. Phase one: a project grant-awarded by Charles University for a three-year project, Computerized Thesaurus of the Czech Language, resulting in a preliminary translated version of the Czech thesaurus and the publication of a sample volume as an output;
4. Phase two: expanded version for publication-moving from translation to original compilation for greater autonomy of the Czech thesaurus and expanding the average of 80 items per entry to 300 using Czech sources; specific rules required for entry structure, the type and order of subentries, etc, to ensure the uniform format of the entries;
5. Compiling the index-to achieve the standard index-length equal to that of the dictionary text, a procedure combining manual and mechanical shortening was devised to abridge the dictionary text;
6. Conclusion. Compilation o f a thesaurus via translation from another language is a possible procedure. Supplementing translation with original compilation based on target-language resources is nevertheless recommended if a truly national thesaurus is to result.
The following is a brief report on how a thesaurus of the Czech language was compiled on the basis of Roget's Thesaurus and the experience gained therefrom.