Spaced Math Repetition

by Derek Smith

Yesterday I was in Denny’s entertaining my nieces when they showed me a math game on Nintendo DS. The game reminded me of an article I read in Wired a while ago about an application called Super Memo. Since 1985 the creator Piotr Wozniak has been using his digital flash card system to train his brain, and using his brain to improve the system. His key insight was to develop an algorithm which monitors your retention accuracy and prompts you to answer questions just as you’re about to forget them, a process called spaced repetition.

Today I went looking for open-source implementations of SuperMemo which allow LaTeX input. I found two: Anki and Mnemosyne. Both of these were available in the Ubuntu Software Center, but versions may be downloaded for all major platforms (Anki provides support for numerous mobile platforms). I found Mnemosyne to be a little more straightforward to use and will be experimenting with it in the coming quarter, but I’m going to explain the basic features of each.

Each uses the abstraction of a deck of flashcards, also allowing you to tag individual cards for additional organization. On my Ubuntu machine, each program rendered LaTeX input without a hitch. You can even alter the preamble to include packages and make your cards as complex as desired (see documentation for details). Anki provides an web account to store your cards and sync among your devices. This seemed helpful, but the web version is not setup to render LaTeX. Editing was more natural in Mnemosyne, as it provides a LaTeX preview. I wasn’t able to figure out how to get Anki to do the same thing so it was more cumbersome when I made a mistake.

I’ve just started playing with these applications today, so I will suggest this more thorough comparison of features. This link suggests that Anki can import the Mnemosyne database, so if you try out both applications maybe start with Mnemosyne. I intend to use this to study both for classes and quals. I am going to input definitions, theorems with proof “ideas” and other problems. Has anybody used such a system before? What kind of success have you had? Did you use it for math and if so, how? I will report back and let you know the outcome of my experiment!


About Derek Smith

Former weather dude and scientific software developer. In the upcoming 2015-16 year I will complete my PhD at UCSB in nonlinear dispersive equations. I enjoy spending time with my two young daughters and running.
This entry was posted in General. Bookmark the permalink.