How the results were createdThe results were created using two separate programs. The rhymes were generated using already existing rhyming code that's described here.
The "sounds like" results were generated by taking the phonentic representation of a word and then adding, deleting or swapping a single phoneme to it and testing that the new phonetic sequence represented a real word. Essentially find all words that had a phonetic edit distance of 1 to the real word.
After combing those 2 results I filtered by frequency. Common words are words that are in the top 40,000 words as ranked by Chapter 14 of Beautiful Data by Peter Norvig. Uncommon words are those ranked between 40,000 and 100,000. And any words that do not appear in the top 100,000 are filtered.