Storage and retrieval of bibliographic records: a comparison of database management system (DBMS) and free text approaches

System designers concerned with bibliographic storage and retrieval systems have, in the past, followed one of three approaches — simple data structures, with the handling of the inherent complexity of bibliographic data embedded in the programs; free text (inverse file) systems of which the ‘online database’ systems may be seen as a specialised extension; and structured database (DBMS) methods, of which some MARC systems are a special subset. The form of storage of the bibliographic data affects not only the efficiency of retrieval by various fields of the record, but also the space used in storage, the cost of maintenance and addition of new material, and especially, the nature of the retrieval language and the flexibility of response to users' needs. The choice of design approach involves study of each proposed application, and guidelines are offered for consideration in matching functional requirements with the varying strengths and weaknesses of the software techniques. Hybrid database structures offering both free text and DBMS features are becoming acceptable, at least for small systems, and the likely paths for future development are noted.