An on‐line system for handling personal data bases on a PDP 11/20 minicomputer

Abstract

Information retrieval is a sequence of selection and discrimination steps, aimed at the production of progressively smaller subsets. A great multitude of documents are prepared for publication by aspiring authors and though editorial control produces some weeding‐out, a high proportion are probably published in one form or another. From this published literature are selected those items to be included in the large indexing and abstracting journals, on the grounds of either quality or subject matter. These publications are then searched by human scanner or computer to select the small minority of bibliographic records which may be of interest to the ultimate user. Then it is the user's turn. He selects from the bibliographic listing a yet smaller number of documents he wishes to view and, finally and occasionally, in a few of the documents he reads, he finds useful information: a datum, an experimental method, the barely discerned inkling of an idea.