Abstract: Information Retrieval System for Well Data in Indiana
Robert F. Blakely, Gerald L. Carpenter, Paul Friesen
A retrieval system designed to operate in a time-sharing environment is divided into four parts. During the first phase, an interactive program operating at a remote terminal without the file interrogates the user and builds up a set of very structured queries. A certain amount of primitive learning is involved, e.g., misspellings are tolerated and statistics are kept on them. The frequency of questions causes the tables to be restructured on the next pass for faster searching. Search statistics also are kept so that the system programmers may analyze the use patterns. After the user has closed his dialogue with the terminal, the system informs the user that his answers will be found on a file at some later time and enters a remote batch mode--the second part of the syste . Files are copied from tape to disk and disk-manager systems are used to speed the search. The interactive program has heavily coded the information so that searching involves no alphabetic well information except for farm name and operator. The user picks up the file for display at his remote console and/or printout at the central site. The third part of the system is an interactive bookkeeping routine that allows the tables for lithologic and other codes to be changed and allows alteration of the misspelling list based on the statistical report. The fourth part of the system is a file-maintenance program that operates in the batch mode to allow the systems-maintenance personnel to create, add to, correct, or delete records in the file.
AAPG Search and Discovery Article #90968©1977 AAPG-SEPM Annual Convention and Exhibition, Washington, DC