Abstract

The amount of information available on both public and private networks continues to grow at a phenomenal rate. This information is contained within a wide variety of objects, including documents, e-mail archives, medical records, manuals, pictures and music. To be of any value, this data must be easily searchable and accessible. Information Retrieval (IR) is concerned with the ability to find and gain access to relevant information. As electronic data repositories continue to proliferate, so too, grows the variety of methods used to locate and access the information contained therein. Similarly, the introduction of innovative retrieval strategies—and the optimization of older strategies—emphasizes the need for an infrastructure capable of measuring and comparing the performance of competing Information Retrieval solutions, but such an environment does not yet exist. The purpose of this research is to develop an infrastructure wherein Information Retrieval solutions may be evaluated and compared. In 1979, an expert in the field believed the need for a system-independent benchmarking utility was long overdue—twenty-five years later, progress in this area has been minimal. Contrastingly, new theories have emerged; new techniques have been introduced; all with the goal of improving retrieval performance. The need for a system-independent analysis of retrieval performance is more critical now.

Degree

College and Department

Ira A. Fulton College of Engineering and Technology; Technology

Rights

http://lib.byu.edu/about/copyright/