Free Text Information Retrieval: An Assessment of Publicly Available Unix-Based Systems

Arisingout of interest inapproaches for content-based image retrieval from theHST image archive, we prototyped free text retrieval systems using accepted proposal abstracts (hence: what the later associated images were meant to deal with). Other possible applications of such free text retrieval systems in observatory operations are described. Weusedwidely available, publicly accessible, (Sun) Unix-based software tools only. Wediscuss these experiments here in terms of client-server implementation, support for phrases in queries, etc. We also review the applicability of the IAU Astronomical Thesaurus, and object name resolvers, in this context. We conclude with a number of recommendations for future work.