A Computer Science View Obtained by Automatic Document Processing

Abstract

Automatic document classification techniques have been widely advocated for the study of various fields of learning, the identification of individual research topics and of influential contributors in a given field, and for information storage and retrieval purposes. Several thousand research documents in computer science are automatically classified in the present study, leading to the generation of a taxonomy which reflects the state of computer science as of 1974. The popularity of various subject areas in the field is assessed, and the clustering characteristics of particular document classes and authors is given.