Very cool idea. I second what mezpub said about saving the audio for import into anki. I think however that attaching sound files to each card would massively slow down card production as it needs to be done 1 by 1. The sweet thing about rikaichan is that you can make cards in bulk (say 100 in half an hour), by using the 's' command and importing to anki from a file. If a system could somehow be set up to import and attach the audio file to each card in bulk that would be awesome. Perhaps rikaichan could be modified to output the name of the audiofile from JDIC as a seperate field, and then anki set up to automatically use this information to attach the sound files in bulk. This is just speculation though and I have no idea how to do it. In addition to this, if anki could be set to display random sentences (taken from japanese wikipedia for example) on the answer side and intergrated with some text-to-speach system that would be the ultimate study system in my opinion.
It would give you:
-massively decreased card production time
-pronunciation for each word
-context for each word
-a random sentence for each time you review the word allowing you to better test your comprehension.
-listening practice
It would give you:
-massively decreased card production time
-pronunciation for each word
-context for each word
-a random sentence for each time you review the word allowing you to better test your comprehension.
-listening practice