Explorations in bibliometric historiography: The (re)emergence of neural networks, 1980–1991
✍ Scribed by Katherine W. McCain
- Publisher
- Wiley (John Wiley & Sons)
- Year
- 2006
- Tongue
- English
- Weight
- 331 KB
- Volume
- 42
- Category
- Article
- ISSN
- 0044-7870
No coin nor oath required. For personal study only.
✦ Synopsis
Artificial Neural Networks (ANN) emerged from exile in the early 1980s as an interdisciplinary field drawing on findings from biology, psychology, computer science (McCain & Whitney, 1994). An earlier generation of research had essentially foundered when Rosenblatt's Perceptron (an early neural network) was shown by Minsky &Papert to suffer from substantial limitations (Jain & Mao, 1996). Beginning around 1984, ANN became a visible presence with dedicated conferences, a professional society, and several journals focusing on ANN theory and applications (McCain & Whitney, 1994).
This poster examines the early history of the current ANN revival-represented in the published literature 1980 -1991-as a platform for examining the usefulness of two different approaches to studying the quantitative history of a research specialty. One approach, developed by the Institute for Scientific Information, focuses on strings of Research Front Specialties -clusters of highly co-cited documents and their source (citing) papers (Small 1977). The other, Algorithmic Historiography, uses Garfield's HistCite software (Garfield, 2004) to plot temporal networks of highly cited papers extracted from the Web of Science and trace the intellectual history of a field.
Defining Research Front Specialties
Research Front Specialities are clusters of heavily co-cited documents and the source documents that create/cite into a given cluster. They were produced annualy for public consumption by ISI until ~1992; ISI kindly provided the data for this study.