Compression Algorithms for the Storage of Bibliographic Information (First of two parts)

BIBLIOGRAPHICAL DATA BASES WHICH HAVE BEEN DEVELOPED IN PC COMPUTERS HAVE BEEN LIMITED REGARDING TOTAL NUMBER OF FICHES AND THEIR PERFORMANCE DUE TO THE SIZE OF THE PLATFORM IN WHICH THOSE DATA BASES ARE DEVELOPED AND INSTALLED. COMMERCIAL PC DATA BASE MANAGEMENT SOFTWARE HAVE BEEN CONSTRUCTED WITH A GENERAL APPROACH, THINKING IN STANDARD APPLICATIONS, AND DO NOT CONSIDER THE PARTICULAR FEATURE OF THE BIBLIOGRAPHIC INFORMATION. THUS, THEY DECREASE IN PERFORMANCE EXPONENTIALLY IN RELATION WITH THE SIZE OF THE DATA BASE. IN THIS FIRST PART OF THE SURVEY, THE PROBLEMS ARE DISCUSSED, AS WELL AS THOSE TYPICAL FEATURES OF THE BIBLIOGRAPHIC INFORMATION REGARDING TO ITS INCLUSION IN COMPUTARIZED DATA BASES. A BIBLIOGRAPHIC DATA COMPRESSION MODEL IS INTRODUCED AS AN ALGORITHM, ALLOWING BETWEEN 40 AND 70 % OF COMPRESSION RATE WITHOUT LOSING INFORMATION QUALITY. IN THE SECOND DOCUMENT, PROCEDURES FOR CREATION AND COMPRESSION OF PRECONSTRUCTED INDEXES WILL BE PRESENTED, AS WELL AS RETRIEVAL FILES FOR WORD FREE-SEARCHING. SOME TECHNIQUES FOR CREATION AND RETRIEVAL OF BOTH ACCESS PATHS WILL BE FULLY DISCUSSED. IN THAT PART THE FINAL CONCLUSION SHOWS THAT DATA BASES WITH SEVERAL HUNDREDS OF THOUSANDS OF RECORDS OWNING SEVERAL MILLIONS OF RETRIEVAL WORDS CAN BE COMPRESSED TO THE AVAILABLE SPACE OF A CD- ROM (650 MB), AND EVEN EXPANDED TO GREATER FIGURES.