The element identified as the main article of a page is passed to a secondary parsing method that further refines the article content by removing extra spaces, punctuation, and transforms into root words. This result is then passed to another function that turns the word list into an array, and passes each word through a series of rules that modify its weight. The exact rules are secret, but it's obvious that the largest effect on weight is the number of occurances. The weighted array is passed back to this page and a script creates a table of all words with a score over 1. Another script is then triggered to read the table and create a line chart.