How to Memorize Faster with the Spaced Repetition Learning Technique | Hacker News Reader

How to Memorize Faster with the Spaced Repetition Learning Technique

productive.fish

177 pointsrossnoel61 comments

The value of SRS goes way beyond memorization. When applied seriously to a field like programming, it enables you to think about the program-design space in a more abstract manner and quickly call to mind and evaluate possibilities. Look up "chunking" as it relates to performance (e.g. in chess).

I've written (and, more recently, made a video) about my 10 years of experience using SRS (via the free tool, Anki) to boost my IT skills.

https://www.jackkinsella.ie/articles/janki-method

Thanks for sharing I might actually start to implement those techniques for further learning.

Thanks! I just watched the video and it looks great, will also read the article later. Also - did you happen to share your (programming related) anki deck? I would love to see that!

Have you by chance tried learning more analytical things like algorithms/mathematical ideas using SRS? I have been trying to automate my interview preparation process using SRS, but it is just mind-blowingly difficult to cardify these things; and cloze deletion doesn’t work well on large algorithms with many pieces.

Note: Please note that I am not advocating that people do it blindly, in fact I myself have been trying to use it as a way to retain once I’ve understood the algos.

This is super interesting, appreciate the post!

I am toying with ideas for my own knowledge base, with integration into SRS for improved retention. I hadn't thought it would go much beyond fact retention, but my hope was to frame facts in such a way that larger pictures could be also retained, analyzed, etc.

It sounds like you've solidified some of the aspirations my idea .. area, so huge thanks for that! This is the first time i've seen mention of SRS going beyond simple facts, and i thought perhaps my idea was a pipe dream.

It seems like there's been a lot of buzz around space repetition on hacker news lately. I've used SRS for learning Japanese with great success, and it would be hard for me to imagine making any progress without it.

The one thing I was missing from most SRS apps was the ability to create note cards, not necessarily in the review schedule, and link them with my cards. There are some concepts like grammar that fit the "note" format, but don't necessarily work as a flashcard. Something like a Zettelkasten + Anki, but with nicer interface like Notion.

So I built an app[0] that does just that with markdown. If that sounds interesting to you, or you've been frustrated with Anki, take look and let me know what you think.

[0] https://mochi.cards/

My experience with the same thing (learning Japanese) has been less than stellar. My rote memory has always been fairly poor, to the extent that I believe my relative academic success compared to my peers has come from being forced to attain a deeper understanding of topics just so I can derive test answers from base principles. I find going through the Anki decks to be seriously draining, especially when encountering the same words I've seen a dozen times and still can't readily recall what the hell they mean. However I've yet to discover a better method for me, since I am not enthusiastic about the idea of immersing myself in Japanese media for large amounts of my free time.

Immersing is incredibly frustrating / challenging when your comprehension is < 90%. It's a real chore.

For me, I've found that having audio for every card is invaluable. Preferably audio from the source where you first encountered the word (for me that's mostly Youtube or Netflix). I'm working on a chrome extension to make this easier. Not only does it make it much easier to remember the word (by a pretty significant margin), but it has done wonders for my pronunciation as well.

I've written a little bit about my process here[0], but it's changed a little since I wrote that.

[0] https://mochi.cards/blog/using-spaced-repetition-to-learn-a-...

If you need a suggestion, Anki is for retaining what you have already learned, not learning, which is what you are supposed to do yourself. In other words, learn to learn. For example,

I heavily rely on mnemonics and I have a field for them, thus for the German word "vertrauen", which stands for "to trust", I have the following mnemonic: ”I trust Ver to heal my wounds”. Ver is a friend doctor of mine.

For Japanese it helps me to have also mnemonics for kanji sounds(ka - cat, ho - hole), then make stories based on them.

Interesting to hear your approach to spaced repetition. I kind of went the other way with Traverse.link, where all 'cards' are interlinked markdown notes, but you have a switch to turn them into flashcards, ie add them to your spaced repetition schedule, and add active recall questions

This looks amazing! I recently (5 days ago) started learning Japanese using Duolingo and have been taking hand written notes. I'm also thinking of creating a digital version using Zettlr[0] which I use daily. I'll also try out your app. If I may ask. What is the tech stack you used for building clients other than mobile as well as for mobile? Thanks.

[0] https://www.zettlr.com/

Whenever a srs thread appears on hn, I hope that the mythical little fish will appear, the famed supermemo user of ten plus years who learned many languages with srs, then disappeared into internet obscurity

Discussion on spaced repetition quickly move on to Anki invariably, but I would like to highlight org-drill https://melpa.org/#/org-drill too. If you are an Emacs user, that is...

That looks interesting! Will definitely check it out. And for those who want to some Anki routine like adding cards via Emacs, see https://yiufung.net/post/anki-org/ (Power up Anki with Emacs).

We are developing an extension for Google Chrome that adds flashcards to browser startpage. This extension is based on a spaced repetition algorithm like Anki, but is much easier to use and understand. Here is the URL of extension: https://chrome.google.com/webstore/detail/studylib-new-tab-s...

Features - You can adjust the number of flashcards and the delay for each session.

- You can choose different types of question (flashcards, multiple choice, written).

- Easyly add new flashcards or import existing sets from other apps.

- Built-in translator in popup bubble (can be swithed off).

Hope this might be of interest to someone ...

That's a very nice idea for casual learning/low number of items to remember.

Edit: Just installed it and the first thing it asks me to login. Sorry, no way. Learn from Momentum how to do it the proper way.

Why do you think so? My language sets, which I study every day, contain over 3000 different flashcards.

Our extension is part of studylib.net platform. Users can learn their sets both in the extension and on the website using PC or phones. This is possible because the user's progress data and sets are stored in one place. So, we need authorization. But you can create account with any email if you don't want to sign in with google or facebook.

Momentum is nice extension, but it has different functionality.

I tried it studylib this weekend and it's much better than expected. Thank you very much for your work and I'm sorry for my first negative reaction.

> for Google Chrome

What about Firefox?

We already have many requests for the Firefox extension. Hopefully we'll add it in January. You can follow https://twitter.com/studylibproject, https://www.reddit.com/r/studylib/ or https://www.facebook.com/studylibplatform to keep up with the news.

i was about to install it but then the permission required are:

- read and change all your data on the websites that you visit...

Just can't accept this level of risk.

For people who already use anki there's a pretty nice chrome extension: https://chrome.google.com/webstore/detail/anki-quick-adder/g...

Here is a recent HN submission also about spaced repetition: https://news.ycombinator.com/item?id=25160204

I personally don't think spaced repetition by itself works very well for a few reasons: -card creation from stuff you read or watch is a huge pain -most people make cards that don't work in real life but don't change them because they assume the algo is good for learning when they're really using it for cramming

I think incremental reading as implemented in supermemo is necessary for SRS to shine: incremental reading works as systematized learning which makes cards in spaced repetition (systematized memorization) much better made

> card creation from stuff you read or watch is a huge pain

Yomichan, sub2srs, Migaku Dictionary's export definitions function, and many other tools simplify the process considerably.

I didn't have language learning in mind at all. If for language learning, completely different. Tools for it are great, I also learn Japanese and mainly use yomichan + voracious and plan to switch into using mpvacious.

Card creation from stuff you read is one of the problems I'm trying to solve with Traverse.link. It has interlinked notes which can be turned into flashcards (adding active recall questions etc.). You can currently paste a link on a note to convert a web article into cards. We're also working on PDF and EPUB import

For illustration, I've turned the linked article into flashcards here https://traverse.link/home/How%20To%20Memorize%20Faster%20Wi...

I figured this out by myself through experience during school/college (I think others have experienced it too) - good to see that validated.

Yeah, it’s basically the same as going through a deck of physical flash cards and after doing the top one, putting it closer to or further from the top of the deck depending on how well you did.

Obviously plus more stat tracking and so forth.

Is this really faster? Or just better for your long term memory. I have always wondered this about spaced repetition because it always feels much slower, but I keep seeing the word “faster” appear in articles that talk about it. Can anyone elaborate or link to some kind of a “speed test”?

I think “more efficient” is probably better phrasing: The amount of calendar time required to reach a given recall level will be longer, but the time spent actively developing that memory will be less and it will fade more slowly over time as compared to massed review (cramming).

In my experience it's faster for learning because you will recall more after a study period if you do the flash cards vs. rereading things a few times. Not an either or thing as well. I would put a lot of my lecture notes in the extra section of the flash cards, then if I want to get more context I can scroll through my cards sorted by created_at in the browser.

Ideal workflow is watching lecture online at 2x speed with Anki add card window open along side lecture window, pausing when it's time to make a card, write the text for the card and use Lightshot screenshot to quickly grab any images or text (or for image occlusion if you're using that, rebind screenshot capture to F4 so you don't need to use your mouse hand to press PrntScrn). Once you're done adding cards, go through them once that day. Or, if you're using a premade deck, watch the lecture, unsuspend cards relevant to the lecture, then go through them right away.

Not sure if you're using it for school or not, but I think the process will work the same either way. Works even better if you're using it along with a memory palace kind of thing like SketchyMicro or Pixorize. Those images will be burned in your brain and your recall will be insane.

if you don't want to silo your cards you might consider NeuraCache (https://neuracache.com/, I'm the creator)

How is a proprietary app with no export option (from what I can tell from the website) no infinitely more siloed than something like Anki?

"no export option" There is no need to export as you can't create content with NeuraCache.

You create cards in markdown or evernote or onenote or roam or ...

NeuraCache serves just as a viewer of flashcards that can be used with many sources.

Okay, I have to admit that I misunderstood you product, and my original comment was unnecessarily harsh.

If the content stays in the original apps I would highlight that on the landing page, as that's not really obvious (at least to me), and goes counter to how most apps operate.

What about the study progress? To a lot of SRS people study progress is almost as precious as the cards themselves.

You have 4 algorithms to choose from currently (SM2 is one of them) https://neuracache.com/spaced/ + I am working on SRS editor that will allow you to define your own intervals.

You can quickly access your decks via tags/notebooks etc

The parent comment has a point, I opened the homepage and searched for the word "data", 0 results found. Perhaps adding something along the lines of "Stay in control of your data with ..." would be a selling point?

Thanks! I will try to make adjustments to the copy

Just a small feedback: I scrolled through 3/4 of the page and still have no clue what is the product. An integration to other apps? A twitter bot? A separate app like Anki? A website version of Anki?

Interesting. Does this also work with the Mac version of MS One Note?

yes, NeuraCache is a mobile app that can sync with OneNote (no matter how you use it)

Isn't this just a fancy name for reviewing?

“Spaced repetition” is a family of algorithms to determine what and when you should review. They’re optimized to minimize the amount of review time necessary to maintain a particular strength of memory over the long term.

What I hate about Anki is lack of a "never ever show this card to me again" button. I could make great use of it when I try to expand my vocabulary and it teaches me basic level words when I only want advanced+ level. This is what stops me from using it.

I would also love to be able to set a custom repetition algorithm and to import cards from a directory of Markdown files.

that would be 'suspend' button

Thank you. I'll try it. I thought it's meant to do what it says - suspend a card temporarily.

I think you want the “delete” button.

I believe I could never find it.

What you want is the bury option, that one will no longer show the card in reviews (but it won't delete it)

Bury (Press = when reviewing) makes Anki not show you a card until the next day. Suspending (! to suspend all cards based on a note [if you have more than 1 cloze on a card, it will stop showing you all of them], @ just to suspend that one card) will make Anki never show you a card until you unsuspend it in the browser.

Hit delete on the keyboard when you are viewing the card.

On the phone the option is there but I believe its nested under options.

> import cards from a directory of Markdown files

Anki supports CSV as an option for import. How about you convert the markdown files to CSV and import it? Recently I used it just for fun to import all the sentences from Faust and put them in a spreadsheet with LibreOffice Calc. I made it separate all the sentences by dots, and added another column with translations of the file thanks to DeepL. Converted it to CSV and it worked out, sentence and translation cards done.

This is exactly what you can do with Traverse.link (well you'd currently have to put your markdown files into a single JSON but I'm working on that). Your markdown notes then become flashcards, but the twist is every note has a "Remember" switch, which you can simply turn off to remove it from your spaced repetition schedule (without losing the note)

Are there "prepackaged" cards for algorithms and other computer science topics?

Possibly, but that is missing the point. A card deck should aid you in building your own model rather than forcing dumb memorization.

A good card is one that highlights what is important for you to memorize. While preparing the card, one should reflect what are these parts and how it fits into the larger whole so as to reflect this in the card. This helps go beyond mere memorization, and apply spaced repetition also to some higher-level concepts.

Example: for a recent exam I took I had to memorize extinguisher types and to which types of fires they can be applied. You can stuff it into a dumb deck ("[Class A]= [common combustibles]"), or you can put it into a hierarchy or turn it into a relationship ("[Class A] extinguishers cannot be applied to [electrical equipment]] because [...]. [Class C] should be used instead.")

For more information, see e.g. https://medschoolinsiders.com/medical-student/anki-flashcard...

The standard advice is to make sure you understand something before you make a card about it, but I've never seen an adequate description of what people mean by "understand" in this context. If a card fails (but follows the other rules), it must mean that you didn't properly understand in the first place, which isn't particularly actionable.

There are two distinct kinds of knowledge: tacit knowledge, the ability to perform some task, and rote knowledge, the ability to recall facts. In most domains, tacit knowledge more useful and harder to come by than rote knowledge. I have come to believe that "understanding" in the SRS context means learning how to do something, how to select an appropriate fire extinguisher in your example. Cards formulated according to those guidelines, then, encode related facts. The idea seems to be that recalling the facts will also reinforce the tacit memory.

An alternative is reinforcing the tacit knowledge directly, which requires a completely different approach: Each card should be a prompt to perform some task you care about, with all relevant context (and possibly extraneous context). This has a reasonable chance of succeeding if you use a premade deck, because all of the information is present. Unfortunately, this isn't how most shared decks are structured.

I've been using a variant on incremental reading for this: I took some books from Project Gutenberg and split the text up into cards and blanked out a word towards the each passage as a Cloze deletion. If I'm successfully following along with the argument that the author is making, then I can usually predict the missing word despite having never read the passage before; otherwise, I need to spend more time with the passage.

Other skills have separate decks: I have a piano deck, for instance, that contains sheet music on the front of the cards with a few bars marked for special attention (or blacked out). I review that one on a tablet while actually at the piano, playing the pieces. I use a photo editor to split up the scan into small parts and throw them all into Anki; it seems to work just fine without the pre-learning that's usually recommended.

This might interest you https://ankiweb.net/shared/info/25173560

Pimsler language audio book works this way.

I found it very effective

Reply on news.ycombinator.com