Linguists use many tools to perform analysis of natural language dialects.

This tools was built from a clustering algorithm - a program that attempts to bucket statistically related words.

From multiple runs of the clustering algorithm, we determine how strongly related words may be, and use graph software to visualize the results.