“The SOM is an algorithm used to visualize and interpret large high-dimensional data sets. Typical applications are visualization of process states or financial results by representing the central dependencies within the data on the map.
The map consists of a regular grid of processing units, “neurons”. A model of some multidimensional observation, eventually a vector consisting of features, is associated with each unit. The map attempts to represent all the available observations with optimal accuracy using a restricted set of models. At the same time the models become ordered on the grid so that similar models are close to each other and dissimilar models far from each other.”
http://www.cis.hut.fi/research/som-research/som.shtml
The map that can be seen through the following link is a great application of the SOM idea made by Andre Skupin:
“Description of Content:
Visualization of the geographic knowledge domain based on more than 22,000 conference abstracts submitted to the Annual Meeting of the Association of American Geographers (1993-2002). Landscape features express the degree of topical focus, with elevated areas corresponding to more well-defined, topical regions and low-lying areas corresponding to a mingling of various topics. Dominant terms are used as labels for topical regions.
Description of Unique Features:
The most unique aspect of this visualization is its combination of intense computation with geographic metaphors and cartographic design considerations. From a computational perspective, the use of a self-organizing map consisting of a large number of neurons (10,000) is fairly unique. The final map presented here aims to explore how far we can go in the design of map-like information visualizations. Its use of a range of label sizes (from very large to very small) on a large-format map and the omission of a legend are aimed at challenging traditional notions of interactivity, by encouraging viewers to vary their distance from the map and instigating discussion.”