I was wondering if there was a way to encode binary data, such as Crypto addresses into a natural language format, in a similar fashion to how what3words.com can encode locations into three easy-to-remember words.
Would it be possible to use something like a recurrent neural network with a deterministic word sampling function to do this? Is there an easier way to go about accomplishing this?
I did some basic calculations and with 171,146 words, it would take a minimum of 15 words to cover a 256-bit-address. Is there a fast way to do this?