Neural networks research in context: A longitudinal journal cocitation analysis of an emerging interdisciplinary field

A cocitation analysis for thirty-six journals and other publications in neural networks research and related disciplines was conducted over three consecutive time periods spanning the years 1990-early 1997, showing an increasing separation between natural sciences/psychology and engineering/neural networks research.

Abstract

A cocitation analysis for thirty-six journals and other publications in neural networks research and related disciplines was conducted over three consecutive time periods spanning the years 1990-early 1997. Cluster analysis and MDS maps identified groupings representing foundation research areas (physics/optics, computer engineering, neuroscience, expert systems & cognition, and perception) along with neural networks and mathematical modeling of neural systems. Principal components analysis demonstrated a similar structure, with several journals and books loading on a majority of the factors. An INDSCAL analysis showed an increasing separation between natural sciences/psychology and engineering/neural networks research from the first time period to the third.