This project aims primarily at creating a natural language interface (NLI) for the Z39.50 interface, an international standard interface. In this way, the project opens a user-friendly access to networked digital libraries and to domain specific information providers on the Internet. It addresses explicitly those users who are not computerliterate (so-called "naive users"). The proposed project will be used to close the gap between methods in natural language processing and access systems on the Internet, which are technologically quite advanced. To the best of our knowledge, there are no NLIs available for the Internet protocol Z39.50 at the moment.
A further aspect of the project is the development of methods for building large computational lexicons. A relevant basic inventory of lexemes that are needed for the practical use of the NLI-Z39.50 will be provided. In a second project phase, the problem of transformation of national and international exchange formats (e.g. MAB, UNIMARC) into suitable presentation formats and the integration of a local relational database in the overall architecture will be treated. In addition to the described components, an NLI to the retrieval language SQL will be developed for the local relational database.