Full-Text Retrieval: Systems and Files

Much of the development in the first 30 years of library automation has been in solving the problem of identifying relevant sources. Automation of the library's card catalog provides a finding tool for the library's collections. The books, journals, films, and other materials located through the catalog still mostly reside in their original form, with no direct connection to the automated finding tool.

Most of the early development in electronic publishing was also aimed solely at identifying information sources. Secondary publishers, notably publishers of indexing/abstracting serials, were the first to provide their resources in electronic form. Throughout the 1970s and much of the 1980s, indexing/abstracting (bibliographic) databases were predominant in the online database world. The first CD-ROM databases for libraries were many of these same bibliographic files. Traditionally (and still) bibliographic databases are the most widely used type of electronic resource in libraries.

Starting in the mid-1980s, due to great increases in disk storage capacities and better document conversion techniques, full texts of certain types of documents became more widely available. In the 1990s,full-text databases (files) are the most rapidly growing type of commercially available database. Better text-retrieval software is leading to more locally created full-text databases as well. Perhaps in this decade we will at last electronically solve the document delivery problem as well as the document location problem.