Counts all places where one match occurs within a short distance of another match.
Arranges the initial matches - up to 49 - in a circle.
Draws lines between matches which occurred close to each other. The thickness and brightness of the lines is proportional to the number of matches.
The halfway point between two matches shows the number of matches (A->B + B->A) if their count is > 4.
Useful for names and places.
I tested the the Jasmine set (combined Jasmine A and Jasmine B dreams from SDDb) there is a strong association between three names; Sandra, Sheryl and Sam. Running a Family search only reveals Sandra and Sheryl as the dreamer's sisters.
I ran a Network search on the results of a Names search combined with an expression which matches 'family identifiers'.
Slightly fainter links run between Mon and Dad, the two sisters and Sam. From this it's an easy guess that Sam is the dreamer's brother.
Also note the strong connections between Uncle and Larry and between Aunt and Christine, suggesting they are a married.
Richard Schweickert, in "Properties of the Organization of Memory for People:
Evidence from Dream Reports" (http://web.ics.purdue.edu/~schweick/Properties.doc) found associations between characters in dream reports that suggest dream accurately reflect waking experience of the dreamer's waking social life. He looked at three dream sets from DreamBank; Merri, Arlie and the Natural Scientist.
Schweickert measured all characters in the sets. I only identify names and then add 'family identifiers' to help determine the roles of some names.
Schweickert also used co-occurrence within a dream as the basis for association between two characters. I use co-occurrence of Names within a 55 character window.
Here's the result for Merri, showing the highest degree names; Dora, mother, Rudy and Father. There is only a faint line between 'sister' (near top of net) and 'Dora' (4 names to the right) because there is only one match were the two are mentioned together.
