Incremental reading and the formulation of items

From SuperMemopedia
Jump to: navigation, search


From: WojtekTM
Country: Poland
Sent: 19th April 2017
Subject: Incremental reading and the formulation of items

Question:

I have got some doubts about using clozes in IR. I have been trying to do IR for the past few weeks and I have encountered discrepancy in the way I create items. On the one side item should contain some important information to be remembered (according to 20 rules), on the other side as I dig the paper and split it into paragraphs and sentences I usually end up with tens of detailed question from just one article. E.g. if I have a paper about Windows, each command definition can generate 5-10 clozes (command name + all options + most important packages names):

dism /image:folder /enable-feature /featurename:feature /packagename:package /source:path /all

I consider these clozes (questions) as irrelevant to the general knowledge. I usually have to rewrite each cloze as an ordinary question multiple times, undermining concept of quick creation of items using Alt-Z and influencing history of repetitions (history of one item contains history of different contents).

The other example of item:

To begin using OAuth2, you need to know a few things about the API you're trying to access:

  • The *url* of the service you want to access.
  • The *auth scope*, which is a string that defines the specific type of access your app is asking for. For instance, the auth scope for read-only access to Google Tasks is View your tasks, while the auth scope for read-write access to Google Tasks is Manage Your Tasks.
  • A *client id* and *client* secret, which are strings that identify your app to the service. You need to obtain these strings directly from the service owner. Google has a self-service system for obtaining client ids and secrets. The article Getting Started with the Tasks API and OAuth 2.0 on Android explains how to use this system to obtain these values for use with the Google Tasks API.

I end up with the following items:

Item1

To begin using OAuth2, you need to know a few things about the API you're trying to access:

  • [...]
  • The auth scope, which is a string that defines the specific type of access your app is asking for. For instance, the auth scope for read-only access to Google Tasks is View your tasks, while the auth scope for read-write access to Google Tasks is Manage Your Tasks.
  • A client id and client secret, which are strings that identify your app to the service. You need to obtain these strings directly from the service owner. Google has a self-service system for obtaining client ids and secrets. The article Getting Started with the Tasks API and OAuth 2.0 on Android explains how to use this system to obtain these values for use with the Google Tasks API.

Answer: URL of service

Item2

To begin using OAuth2, you need to know a few things about the API you're trying to access:

  • The url of the service you want to access.
  • The auth scope, [... explain].
  • A client id and client secret, which are strings that identify your app to the service. You need to obtain these strings directly from the service owner.

Answer: which is a string that defines the specific type of access your app is asking for

Item3

To begin using OAuth2, you need to know a few things about the API you're trying to access:

  • The url of the service you want to access.
  • The auth scope, which is a string that defines the specific type of access your app is asking for. For instance, the auth scope for read-only access to Google Tasks is View your tasks, while the auth scope for read-write access to Google Tasks is Manage Your Tasks.
  • A client id and client secret, [... explain purpose]. You need to obtain these strings directly from the service owner. Google has a self-service system for obtaining client ids and secrets. The article Getting Started with the Tasks API and OAuth 2.0 on Android explains how to use this system to obtain these values for use with the Google Tasks API.

Answer: which are strings that identify your app to the service

I have found my own items too wordy and now I am going to shorten it substantially. Maybe should I reformulate the last question in the following way, leaving the deleted text in the source Topic that should not be Dismissed from the learning process?

Name three things you have to know to use OAuth2:

  • The url of the API.
  • The auth scope.
  • A client id and client secret, [... explain purpose].

Answer: which are strings that identify your app to the service


Could you give me some hints how to improve it?

Answer:


Another User's Thoughts

(Apologies for the long text -- this is something I've dealt with as well and had to learn how to deal with it)

Speaking strictly as a fellow user who also uses SuperMemo to learn some technical subjects, I've encountered much the same issue. And in my opinion you are on the right track. I've learned to work with it much the same way you have with your last example: reformulate and compress in my own words. When I started using SuperMemo a little over a year ago I would naively cloze delete rather long passages. That worked okay for a while, but now that some of those cards are coming up for review after many months I found that there was very little retention of some of them because I was focused on clozing the passages quickly and moving on instead of capturing the 'essence' of the material instead. So I've learned to reformulate those problematic cards when I find them, to shorten the phrases or rewrite the sentences, or to simply delete the card entirely if it is no longer of interest.

One of the biggest leaps in my use of SuperMemo came when I began 'ruthlessly' reformulating items as I encountered problems like this. Being willing to simply rewrite what is there, even if it "changes" the original author's words, is not only ideal but is 'necessary' for most uses, since in many cases we want to learn the concepts and not the specific words. If I find a card difficult I will even consider completely rewriting it and generating all new cloze deletions that make more sense. Or rewriting it as a "why" question instead of a fact-regurgitation question, if it turns out I recall the concept but not the detail and the concept turned out to be more important in the long run.

For specifics, your last example is much better, but I would consider creating the cards as follows: (again this is all in my opinion)

Card 1:

Name three things you have to know to use OAuth2:

  • The [...] of the API.
  • The auth scope.
  • A client id and client secret

Card 2:

Name three things you have to know to use OAuth2:

  • The url of the [...].
  • The auth scope.
  • A client id and client secret

Card 3:

Name three things you have to know to use OAuth2:

  • The url of the API.
  • The [...] scope.
  • A client id and client secret

Card 4:

Name three things you have to know to use OAuth2:

  • The url of the API.
  • The auth [...].
  • A client id and client secret

Card 5:

Name three things you have to know to use OAuth2:

  • The url of the API.
  • The auth scope.
  • A client [...] and client secret

Card 6:

Name three things you have to know to use OAuth2:

  • The url of the API.
  • The auth scope.
  • A client id and client [...]

And then perhaps additional cards like:

'In OAuth2, the auth [...] is a string that defines the specific type of access your app is asking for.'

And then in the answer have the additional info.

Yes I actually would consider making all those cards -- if I wanted to memorize OAuth2 in that detail. (Microsoft kept changing it confusingly a few years back, something to consider....)

And though it goes against the traditional SuperMemo teachings I might also consider making a "double cloze" for the last one by copying and pasting the [...] markup in the question and adding the answer to the answer field. I don't do this a lot but when I do I will typically do it like this:

  • A client [...] and client [...]

Answer: id secret

Using 5-8 spaces between answers in this case really helps -- I write the answers very quickly and it is easy to distinguish them. I also found that adding a double space around variable names, values etc helps call them out from surrounding text without requiring any special markup and plus its as fast as hitting the spacebar a couple extra times. Here's an example from earlier today, not a cloze deletion but it shows the spacing. The top is the question, bottom is the answer, and the image displays during the answer phase. https://i.imgur.com/VNTsiEi.png (the image is heavily modified from page 7 here: http://www.dspguide.com/CH24.PDF)

The point is not to be afraid to create a lot of cloze deletions. In my opinion. Elsewhere I had asked a question about concerns I had that a particular cloze card was "too easy" and wondered how to make the card take more effort, and found out I was wrong -- easy cards are a sign you remembered the material well. So that changed how I approach cloze deletions, and most tend to be very short, usually one word or a very short phrase.

Here's an example of a set of cloze cards I made after reading Claude Shannon's 1949 paper Communication in the Presence of Noise. He had a remarkable way of viewing signals as geometric points in space and I wanted to remember that concept, so I summarized his geometric view into one card with a lot of cloze deletions. This card is very wordy (far more than 99% I have) but I chose to leave it there because I found the concept so interesting that I want to be able to reread it anytime I come across the concept in a later repetition, in case I don't recall the details. https://i.imgur.com/Rf5AsOo.png

That one card has 27 cloze deletion children! And it seems to work very well for my needs. In fact I created a whole folder in the knowledge tree just for that paper, in this image you can see that one card highlighted with its 27 children, and quite a few other cards with concepts from the paper as well. https://i.imgur.com/WWwpLlo.png

Embrace the cloze!

Here's a final (finally!) image showing an example of capturing concepts. In this case I wanted to make sure to remember the relationship between these three items. In this case a cloze deletion doesn't work, so I made a card calling for a "plain English" aka "whiteboard" explanation of the concept: https://i.imgur.com/DVs2xqu.png (and I still mix up jump tables and dispatch tables)

So in summary I would suggest the following:

  • Try to embrace having many small cloze deletions and see how that works for you.
  • Be sure to create cards that capture the overall concept you want to remember, in addition to specific cloze deletion cards.
  • Remember to attack critical info from different angles -- cloze delete it, then create some scenario questions that call for that answer, and include some "why?" questions as well. (I sometimes ask AND WHY??? at the end of a card that asks for a simple answer, a 2-in-1 card and it works pretty well too)

Hope all that helps in some way.