Hacker News Reader

Comment by moeris | Hacker News Reader

There's also the major system, which is probably better suited to English.

https://en.wikipedia.org/wiki/Mnemonic_major_system?wprov=sf...

Of all the memory systems I learned from Harry Lorayne's books, this is the one consistent one I use. Not for long strings of numerals, but things like PIN codes and phone numbers. It's well worth it.

I once tried to write a script that would generate the most efficient encoding of an arbitrary length number using the Mnemonic Major System. Definitely an interesting problem!

Ha, do you have that public anywhere? That seems like it would be fun to play with.

Edit: Here's a really well executed online tool that someone else made: https://pinfruit.com/

------

Went digging through my old files, turns out this was almost 10 years ago!.. and looks like I left it unfinished.

I was using nltk to convert cmudict words into MM sequences using phenomes.

Then I was basically brute forcing all the different ways you could slice the given sequences into sub sequences. Then the idea would be to try and identify the "best" sequence.

You'd want to restrict it to using relatively common words, preferably nouns. Really you'd want to avoid any words that are ambiguous or could be easily confused.

Even better would be able to find grammatically correct sentences that fit the sequence.. but I think it's unlikely you'd have any hits.

Anyway.. this is the broken/unfinished code I found in the depths of my hard-drive: https://github.com/johtso/mnemonic_major

Reply on news.ycombinator.com