This project continues research into data structures which allow fast searching. A major area of interest involves data structures for rapidly scanning set of elements to find the subset that lies in a particular geometric region, and to thereby develop a better understanding of "range query" costs (from the vantage points of current software needs, potential geometric query objects, a dynamic setting, a static setting, tradeoffs between retrieval time and memory space.) A second goal is to continue research into how to efficiency implement a non-procedu- ral set-theoretic database query language, similar to the rela- tional calculus or SETL, in a main memory database. Other areas of study will include possible examinations of dynamimc perfect hashing, compact representations of sorted lists and communication delays inherent in parallel processing. The importance of this research is that geometric region searching has applications into a wide range of area where the data can be though of as mapped out in space - VLSI circuit design, graphic-design interfaces, and census or statistical database applications based on geographical regions. Efficient retrieval, using advanced technology such as very large main memories for computer, is critical to developing these applications.
