An automatic abstracting program employing stylo-statistical techniques and hierarchical data indexing

This paper reports the progress being made at RCA in the field of Automatic Abstracting. It includes(i) A general formulation of the problem, (ii) The use of stylo-statistics and hierarchical indexing to extract words associated in concept, (iii) Minimum procedures to reduce redundancy within sentences and (iv) A description of the computer program characteristics.