Retrieve: An Engineering Tool for Searching Remote Sensing and Environmental Engineering Databases

The design and development of a semi-automatic information retrieval system which features manual indexing, and an inverted file structure is presented. The system requires manual indexing done by an expert in the subject field to ensure high-precision searching. High-recall is achieved through the implementation of the inverted file. The system provides an interactive environment, a thesaurus for normalization of the indexing language, ranking of retrieved documents, and flexible output specifications. The purpose of this thesis is to present the design and development of in-house search-aid software for small document collections intended for Remote Sensing and Environmental Engineering users.