This type of analysis analyzes minimal correlations between nominal columns in the same table and gives the result in a chart.
In the chart, each column will be represented by a node that has a given color. The correlations between the nominal values are represented by lines. The thicker the line is, the weaker the association is. Thicker lines can identify problems or correlations that need special attention. However, you can always inverse edge weight, that is give larger edge thickness to higher correlation, by selecting the Inverse Edge Weight check box below the nominal correlation chart.
The correlations in the chart are always pairwise correlations: show associations between pairs of columns.