News for hackers. Stuff that matters.
Re:Incremental Learning – A continuous learning approach (2018) by keiferski on Tuesday February 12, 2019 @08:13AM (#19143168)
I have been personally experimenting with what I call "drip learning" - essentially, I am trying to learn a large amount of non-urgent information very slowly, piece by piece, over a very long timeline. For example, learning all the letters of the Greek (or Hebrew, or Russian, or Armenian) alphabet. If I sit down and attempt to learn them all it once, it quickly becomes overwhelming and retention is low. But, if you simply learn one new letter per day, it's very manageable - and you'll learn the entire alphabet in a month.

Combined with spaced repetition (Anki) to reinforce the memory, I think it's extremely powerful. Anecdotally, it works well - I can recall all of these items on demand, months later.

The next step (which I haven't quite figured out) is to multiply this across multiple domains. I.e. every day, learn 1 French phrase, 1 Spanish phrase, 1 German phrase, and so on. In my (nonscientific) experience, this is more effective than focusing on a single topic.

Re:Incremental Learning – A continuous learning approach (2018) by misiti3780 on Tuesday February 12, 2019 @09:39AM (#19143768)
I like the idea but I think with languages it's better to stick with one till you a certain level (maybe b1/b2) then move on.

If you are going to Italy soon you can memorize "non lo so" to say "I dont know" very effectively. But if you are actually trying to learn and retain "some" Italian, you probably need to know more like "lo" is a direct objective pronoun for him/her/it and "so" is the verb essere conjugated in the present tense for io.

Re:Incremental Learning – A continuous learning approach (2018) by xthestreams on Tuesday February 12, 2019 @02:33PM (#19146701)
Little correction: "so" is the conjugated verb "sapere" (to know)
Re:Incremental Learning – A continuous learning approach (2018) by misiti3780 on Tuesday February 12, 2019 @03:01PM (#19146959)
yep - thanks!
Re:Incremental Learning – A continuous learning approach (2018) by marviel on Tuesday February 12, 2019 @08:44AM (#19143346)
You may not be an Android user as I am, but if you are --- im excited for the V2 scheduler in Ankidroid, which will allow you to pull in cards from multiple subdecks when reviewing a parent deck. The desktop client already has it.

If that is added, or you're already just using the desktop client, then IIUC the feature you seek (1 new concept, per category, per day) could be acheived by setting the "new card" number on each deck to "1", provided you've separated each of the concepts into subdecks of a main deck.

Re:Incremental Learning – A continuous learning approach (2018) by dotancohen on Tuesday February 12, 2019 @09:43AM (#19143801)
I could have posted this exact post. I use Anki to slowly study subjects that I'm interested in, including foreign alphabets. Myself and my oldest daughter are doing Greek right now.
Re:Incremental Learning – A continuous learning approach (2018) by xkfm on Tuesday February 12, 2019 @10:21AM (#19144138)
You can do this in Supermemo quite easily. It does have interleaving (mixing multiple "decks") by default, although you can do reviews from one deck if you want to.

You'd be losing FOSS, Linux, and IMHO my favorite feature, the mobile support by switching to Supermemo.

When I was using Anki in school I just used multiple decks. Usually one per class.

Re:Incremental Learning – A continuous learning approach (2018) by sigi45 on Tuesday February 12, 2019 @10:04AM (#19143999)
The most effective way to learn something is actually to sit down and learn it.

There is a significant time / resource difference between one word a day and constant study time where you do learn way more than 1 new vocabular per day in avg.

Re:Incremental Learning – A continuous learning approach (2018) by ddelt on Tuesday February 12, 2019 @10:42AM (#19144375)
Right, but I believe his method was trying to optimize retention.

Depending on well you retain information through a block of constant study time vs. a much smaller, long-view approach to the same amount of material, could be the determining factor on which strategy you choose.