New Releases by B. Doyle

The crucial question of the quality of automatic classification is treated at considerable length, and empirical data are introduced to support the hypothesis that classification quality improves as more information about each document is used for input to the classification program. Six non-judgmental criteria are used in testing the hypothesis for 100 keyboard lists (each list representing a document) for a series of computer runs in which the number of words per document is increased progressively from 12 to 36. Four of the six criteria indicate the hypothesis holds, and two point to no effect. Finally, the future of automatic classification and some of the practical problems to be faced are outlined.

Will there be a breakthrough in the field of information retrieval. One authority in that field has said, ''''No.'''' This paper adopts the opposite viewpoint, and speculates on what the elements of such a breakthrough might be if it were to occur. Several breakthroughs in other fields are scrutinized in order to highlight the factors which characterize and energize sudden expansions of new technologies. These factors, plus some factors specific to the field of information retrieval, are then extrapolated into a ''''plot for a breakthrough.''''

It is argued that the use of ''relevance to a search request'' as a criterion of what a system retrieves is, in effect, a suboptimization on the machine side of the man-machine interface. Also that the searcher needs an efficient exploratory system rather than a request-implementing system. (Author).