FAQ: Knowledge Structuring

Do you have a question about writing simple and effective collections for SuperMemo? Write to Dr Wozniak. If you would like to have your items evaluated, please send 3-5 for review. Do not send whole collections in an e-mail attachment

See also: 

(Mike Condron, USA, Dec 10, 2000)
Question:
I am finding that anatomy is a subject that consists largely of lists. The branches of the axillary artery are... the layers of the abdominal fascia are ... the contents of the posterior triangle of the neck are ... and so on. What is the best way to learn such lists?
Answer:
Enumerations should be discouraged. These always result in recall problems. You should rather use multiple graphic deletion tests. You can, for example, use a picture of the branches of the axillary artery, occlude or point to a single branch in the picture and ask: Which branch is pointed to in the picture? You will have as many items as there are branches of the artery.
See also:

(John Meritt, UK, Nov 26, 1999)
Question:
How can I best learn programming with SuperMemo?
Answer:

  1. You could start from reading SuperMemo as a new tool increasing the productivity of a programmer. Please review the examples of simple items related to programmer's ABC
  2. You will probably have the documentation of the programming language available in electronic form. You could process it using incremental reading, i.e. extract manageable portions of knowledge and reformulate them using cloze deletions. This will save you lots of time on retyping; however, you must be sure you make a good selection and memorize only the most important facts and rules
  3. Do not forget about combining practical sessions with your SuperMemo repetitions. Only practical training will make you truly understand the importance of individual elements of knowledge

(John Meritt, UK, Nov 26, 1999)
Question:
How can I best learn spelling with SuperMemo?
Answer:
You should use/create a template in which the answer will be a Spell-Pad (i.e. text input component). In the question part you should ask about the word that is to be spelled. Because many spelling problems come from the use of double letters (e.g. traveling vs. travelling), you do not need to define the word. It is enough you ask to choose a correct variant. For example, your question might look like this: traveling/travelling or better yet trave(l/ll)ing.
It is very important to focus on one problem at a time. This is why instead of o(c/cc)a(s/ss)ion, you should create two items:

  1. o(c/cc)asion
  2. occa(s/ss)ion

Occasionally, you can make an exception to this rule. For example, you might ask Mi(s/ss)i(s/ss)i(p/pp)i as a request to spell Mississippi. In this case, it is easy to remember that all questionable consonants in this word must appear in double. Once you realize that, you may never experience problems with recalling how to spell Mississippi

(Luis Gustavo da Silva, Brazil, Nov 9, 1999)
Question:
What is your view on the formulation of items in which the question lists the symptoms of a disease and the answer provides the name of the disease?
Answer:
This sound like knowledge that should not cause much trouble in learning:

  1. symptom enumeration concerns the question and as such should not be a problem (unlike enumerations in answers)
  2. this is based on active recall. The list of symptoms will be reviewed passively but the answer will have to be recalled
  3. simplicity of wording may be critical for success here!
  4. the worst problem may arise from the confusion of similar symptom sets but ... resolving such minor differences might be an important skill of a physician. In other words, there is no point in avoiding difficulty in cases of knowledge that has to be mastered one way or another
  5. the best way to figure out the effectiveness of formulation is to use the leech catcher to periodically sift through the hardest items (to find out which might be reformulated)

(Jim Ivy, USA, June 4, 1997)
Question:
What is the difference between a topic and an item?
Answer:
Topics are used to store articles or other content that presents knowledge, while items are used to test knowledge by means of repetitions (usually they have the question-and-answer structure or are cloze deletions). See: Topics vs. items

You can learn programming language syntax
(Noah Chanin, Friday, March 11, 2005 1:33 AM)
Question:
I want to memorize all of the valid forms of SQL. If you have some ideas about effective techniques to employ for learning syntax, it would be appreciated
Answer:
Before you begin, you should be aware of the cost-benefit balance. In most cases, syntax requires no SuperMemo. Simply put, your daily or regular use of the language will ensure this knowledge is easily retained without repetition. However, if you are only a beginner, or you plan to know the language without using it much on a regular basis, incremental reading is always the best option. You can import your entire programming language manual, and process it in proportion to priority of individual subjects. In the end, you will arrive at simple cloze deletions as quoted below. It is worth remembering though, that there is not much benefit in memorizing details of syntax until you realize at some point that not remembering a particular piece of information is a stumbling block in your further progress. Knowledge of a single programming language is usually vast enough to take years to master to the last detail, even with the employment of the best techniques of incremental reading. This is why a good principle is: "memorize only in need" and process the rest passively to reduce time costs. Here are some examples of clozes to which you arrive in the end:

Q: mysql: To rename a table named table_name to new_table_name, use:
Q: mysql> [...] TABLE table_name TO new_table_name;
A: RENAME

Q: mysql: To rename a table named table_name to new_table_name, use:
Q: mysql> RENAME [...] table_name TO new_table_name;
A: TABLE

Q: mysql: To rename a table named table_name to new_table_name, use:
Q: mysql> RENAME TABLE [...] TO new_table_name;
A: table_name

Q: mysql: To rename a table named table_name to new_table_name, use:
Q: mysql> RENAME TABLE table_name [...] new_table_name;
A: TO

Q: mysql: To rename a table named table_name to new_table_name, use:
Q: mysql> RENAME TABLE table_name TO [...];
A: new_table_name

(Noel Clary, USA, Sep 6, 1998)

Item reviewed:

Q: a rod-and-tube element temperature sensor consists of:
A: a high expansion metal tube containing a low expansion rod. The rod & tube are attached on one end. The tube changes length with changes in temperature, causing the free end of the rod to move

Suggestions:

This is a typical case of combining a number of items in one with detriment to the ability to recall the combined item. The suggestion here is to split the item into a number of simpler items that reproduce the same information in student's memory:

Q: What are the two parts of a rod-and-tube temperature sensor?
A: rod and tube

Q: What is the expandability of the tube in rod-and-tube sensor?
A: high

Q: What is the expandability of the rod in rod-and-tube sensor?
A: low

Q: How is temperature indicated in the rod-and-tube sensor?
A: tube moves relative to the rod

Q: Where are rod and tube connected?

etc. etc.

Keep only one deletion in cloze deletions
(Noel Clary, USA, Sep 6, 1998)

Item reviewed:

Q: Step 5 HVAC duct design: Size ducts by the selected design method. Calculate system ..., then select ...

Suggestions:

Cloze deletions (i.e. questions with blanks) are generally a good learning tool; however, in most cases it is better to keep a single deletion per question

(Deron Isaac, USA, May 21, 1997)
Question:
How can I edit texts of items during repetitions without backing out of the test mode?
Answer:
Choose Q to edit the question, A to edit the answer, or E to edit all text components. You can also edit all properties of all components by using the component menu available with the right button click on a component in question

(John Gibney, Australia, Sep 16, 1998)
Question:
In your materials you write that users should avoid memorizing sets (e.g. countries of Europe) or long sequences (e.g. the alphabet). What if I want to remember the sequence of a form of Tai Chi?
Answer:
Let us consider an example in which you want to memorize the entire sequence of letters in the alphabet. It won't be very effective if you use the following item:

Q: What is the sequence of letters in the alphabet?
A: A,B,C,D,E,F,G,H,I,J,K,L,M,N,O,P,Q,R,S,T,U,V,W,X,Y,Z

You will notice that you frequently stumble on parts of the sequence and need to stop repetitions just to exercise the entire sequence in the traditional way (like we all learn poems by rote).

However, you can approach this in a way that guarantees quick effects:

Q: What is the sequence of letters in the alphabet between A and E?
A: A,B,C,D,E

Q: What is the sequence of letters in the alphabet between D and H?
A: D,E,F,G,H

Q: What is the sequence of letters in the alphabet between G and K?
A: G,H,I,J,K

etc. etc.

After 2-3 weeks of repetitions, you may take on an extra task of recalling the whole sequence after each repetition of these simplified items. This will make sure you can recite the entire alphabet quickly. You will also frequently rehearse that parts of the sequence that are harder for your memory (e.g. V,W,X,Y,Z) as opposed to those that are much simpler (e.g. A,B,C,D,E)

Question:
What is the purpose of the option Search : Categories?
Answer:
SuperMemo introduces the concept of a category to help you keep items related to different subjects in different branches of the knowledge tree (contents window). It also makes it possible to give different item groups different appearance (e.g. size of text components, screen layout, font, color, and many more). Read more: Using categories

(Anatolyi Lipatov, Ukraine, Jul 12, 1998)
Question:
I am using Advanced English to enhance my English and business English. Now I am registering for CFA examination (that is Chartered Financial Analyst program of Association of Investment Management and Research). There are several organizations developing and distributing methodological stuff for preparing to the exam. A lot of things should be memorized for passing the exam. What do you think the best way to fit SuperMemo for memorization is and what approach should I use to prepare my own knowledge base for memorizing the material. Is special programming knowledge needed for it?
Answer:
No specialist knowledge is needed to prepare simple knowledge collections in SuperMemo. With Alt+A (i.e. Add new) you get the core functionality! All advanced editing options can be worked around by an appropriate questions-and-answer approach. Perhaps it would be useful yet to learn how to add images to your items (see help for details). To learn more about effectively structuring knowledge in SuperMemo you might want to read 20 rules of formulating knowledge in learning and Knowledge Structuring and Representation; however, nothing works better as learning on one's own mistakes in formulating knowledge for learning with SuperMemo

It is recommended that you keep all your knowledge in one collection
(David Mckenzie, New Zealand, Apr 8, 1998)
Question:
Is there any point in keeping collections separate?
Answer:
No. Once you master contents categories and templates, and Postpone, there is no point. You gain global search, global registries, global repetitions, global optimization, etc. This would not be advisable back in SuperMemo 7 as item difficulty measure (E-factor) was dependent on the average difficulty of items in the collection. Presently, the item difficulty measure (A-Factor, or absolute difficulty factor) is absolute and does not depend on the context in which an item is placed (see: SuperMemo Algorithm). Only the length of the first interval will significantly be affected by the average difficulty of items in the collection. However, this shall not bear dramatically on the speed of learning. Especially that variable forgetting index for individual items makes it possible to set different first intervals for whole contents categories or branches of the knowledge tree

Question:
What should I use Duplicate for? What for do I need the same item in the same collection?
Answer:
You can duplicate an item, if you want to add to the another item which is only slightly different. This way you can spare some time by reediting the old item instead of typing in the new one

In SuperMemo we do not create sequences of items for review
(Marc de Ruiter, China, Apr 5, 2000)

Question:
It is possible to have SuperMemo do follow-up questions like this:

  1. Question
  2. Answer, which is also a question
  3. Answer

I am learning Chinese Characters, so first I have a character which is the Question, then I have to give the answer which is the pronunciation of that character, but then I have to give the meaning in English. So basically there are three things: character - pinyin - translation
Answer:
This is a case where you would like to learn three associated things A-B-C, where A is your question, B is the follow up and C is the follow up to B. It is never a good idea to learn more than one thing in a single item! SuperMemo needs to separately understand your difficulties with linking A and B, B and C, as well as A and C. Building a test according to your suggestions is possible, yet it would be better to use cloze deletions here. For example:

Parent template item:

A B C

Items generated with cloze deletion (e.g. using Reading : Remember cloze):

Item 1:
Q: ... B C
A: A

Item 2:
Q: A ... C
A: B

Item 3:
Q: A B ...
A: C

It is important to know that Item 1 above may make you fail to answer with A to the question C if you only learn to answer Item 1 by understanding the association of B with A. In such cases, you will need even more work by formulating items: A-B (where A is the question and B is the answer), A-C, B-A, B-C, C-A, and C-B. Although you will get six items instead of one, you knowledge is likely to be more solid and you may actually spend less time on repetitions of those multiple items than on repetitions of the conglomerate A-B-C item

See also: Sequential review may be inefficient

(Patrick Mon, Jan 14, 2001)
Question:
Can SuperMemo be used to memorize a hundred digits of PI? What about memorizing very complex formulas or foreign language characters? The 20 rules page did not include those things
Answer:
Memorizing PI falls into the domain of mnemonic techniques. You could try to use SuperMemo to memorize PI as is but it would be tremendously inefficient. On the other hand, if you use mnemonic techniques to memorize things, you should use SuperMemo to handle the periodic review. Despite the claims of many mnemonists, mnemonic techniques will not make you remember for ever. In other words, to memorize PI, use both mnemonic techniques and SuperMemo. For memorizing numbers, the best technique is called the peg-list method. A peg-list is a list of 10 or 100 pictures associated with numbers 0..9 or 0..99 respectively. Memorizing PI is then equivalent to building a list of pictures that correspond with the number. The rule of the thumb in reference to mnemonic techniques is: use them on material that is particularly difficult to remember (e.g. numbers, sets, lists, etc.); do not use them on material that stick well to your memory in SuperMemo. To learn more about mnemonics, see some of the links below:

Question-answer swapping will usually involve some reediting
(Patrik Nilsson, Thursday, July 26, 2001 5:41 AM)
Question:
SuperMemo website claims that swapping questions with answers is rarely applicable beyond word-pair learning. I disagree. If you show a picture of a car to a little child and ask for what's at the picture, next time you ask the child what a car looks like and want him to draw it or explain it by words
Answer:
You are right about swapping the roles of stimulus and response. However, in practice, optimum knowledge representation will enforce a degree of editing. In your presented example, we would rarely use name-picture pairs (unless the whole collection is uniform and leaves no doubt as to the question format). Optimally you will need a verbal clue: What object is presented in the picture? and then How does a car look like? Those minor editing enhancement substantially speed up the learning process

The best way for providing context is to use labels inside your questions
(Steve Brown, Tuesday, August 21, 2001 4:01 PM)
Question:
Is there a way to tell what category a Q/A pair belongs to. The same question may have a different answer. For example, "what is a strike" could have 2 different answers depending whether I was in the "bowling" category or the "union" category
Answer:
Instead of using categories for such a purpose, you could accomplish the same with domain labels inserted before the question. This is the recommended method which immediately evokes the appropriate context. For example: sport: What is a strike? or econ: What is a strike?, etc. This is how most collections in SuperMemo Library are formulated

Combining multiple pieces of information in one item is not recommended (#6101)
(Telepolis, Spain, Thursday, December 20, 2001 1:57 PM)
Question:
What would happen if you presented the pronunciation sound at the same time as its correspondent word? If you are not looking at the screen you are improving the comprehension, and if you are looking, you can check the spelling too. The associative learning will increase
Answer:
Ideally, you should adjust the mode of repetition to your goal. For that purpose it is better to separate learning the pronunciation, from learning the spelling, and from learning the semantics or synonyms. This means that you can create several items for the same word. Experience shows that this is the most efficient method. This is how Advanced English was designed. However, it is rather impractical to produce all combinations of items for all words and synonyms. For that reason, Advanced English includes the pronunciation branch that includes only the words that are hardest to pronounce. Similarly, the spelling branch lists only words that cause most problems with spelling. Bombarding the brain with many stimuli at the same time may produce ambiguous stimuli and you will not always learn that what you really want to learn. Some important aspects of information can be lost

Incremental reading requires some experience
Question:
I do not know how to tackle this text in incremental reading. Any hints?

After the discovery of Pluto, it was quickly determined that Pluto was too small to account for the discrepancies in the orbits of the other planets. The search for Planet X continued but nothing was found. Nor is it likely that it ever will be: the discrepancies vanish if the mass of Neptune determined from the Voyager 2 encounter with Neptune is used. There is no tenth planet

Answer:
Here are some exemplary processing stages. Yours might be different. In the end, you can convert the cloze deletions into more direct and well-formulated questions-and-answers:

Extract 1: Pluto is too small to account for the discrepancies in the orbits of the other planets

  1. [...](planet) is too small to account for the discrepancies in the orbits of the other planets
  2. Pluto is too [...] to account for the discrepancies in the orbits of the other planets
  3. Pluto is too small to account for the [...] of the other planets
  4. Pluto is too small to account for the discrepancies in the orbits of [...]

Extract 2: Pluto was too small to account for the discrepancies in the orbits of the other planets. The search for Planet X continued but nothing was found

  1. Pluto was too small. The search for Planet X continued and [...] was found

Extract 3: Pluto was too small to account for the discrepancies in the orbits of the other planets. The discrepancies vanish if the mass of Neptune determined from the Voyager 2 encounter with Neptune is used

  1. Pluto was too small. The discrepancies [...] if the mass of Neptune determined from the Voyager 2 encounter is used
  2. Pluto was too small. The discrepancies vanish if the new [...] of Neptune is used
  3. Pluto was too small. The discrepancies vanish if the mass of [...] determined by the Voyager 2 is used
  4. Pluto was too small. The discrepancies vanish if the mass of Neptune determined from the [...] encounter with Neptune is used
  5. Pluto was too small. The discrepancies vanish if the mass determined from the Voyager 2 encounter with [...] is used

Extract 4: There is no tenth planet

  1. There [is/isn't] the tenth planet
  2. There is no [...]th planet

Developer's Guide for authors of SuperMemo collections
(anguskwong, Wednesday, August 28, 2002 8:01 PM)
Question:
I would like to develop a knowledge collection similar to 100 English Spellings for Kids with sound, spell pads, etc. It seems I can not do all these as a normal user. Do I need Developer's Guide to do that?
Answer:
You can develop such collections with SuperMemo. It is fully functional if you turn on Full access and Professional level. Please have a look at these:

Does minimum information principle hold in all cases?
(Simon J., Apr 08, 2004, 08:59:51)
Question:
My wife is using Advanced English and has now gone through about 9000 words. However, each time there is a big gap (say 1 month), she will always forget some items (e.g. amenable=responsive/susceptible). Her forgetting index is set at 7%. It seems the word is too abstract and floating in space. Especially when both of the words were previously unknown. I wonder if the 'minimum information principle' holds true for everyone
Answer:
Not all items in Advanced English are structured ideally. We try to locate such cases and improve them in future upgrades. This is what differentiates Advanced English from a simple dictionary. However, it is not always possible to adapt this material for everyone's needs and some individual intervention is always necessary. You are right that it is unacceptable to learn a word pair if both words are not understood. Adding translations or one's own examples in such cases is very important. You could for example start with finding exemplary sentences where "amenable" is used and add them to your Advanced English. You might add then two additional items like these:

Q: Middle East is not [...](responsive) to improvements along American lines
A: amenable
Q: Some things are not [...](open) to the approaches of science
A: amenable
Naturally, for passive recognition you need an item too. This could then be:
Q: amenable (adj)(e.g. some things are not amenable to science)
A: open/susceptible
As for the minimum information principle, you should remember that it does not refer to the number of words or characters in the question. It refers to the information that needs to be stored in the brain. For that reasons, cloze deletions built of exemplary sentences (as indicated above), are FAR easier to remember. Even though they are wordy, they produce a very simple association link in memory. They are highly recommended. As for the forgetting index, you should rather keep a higher forgetting index for the entire collection and lower it only for the most important items. Do not use the forgetting index as a remedy against imperfect structure of knowledge! Modify your items first and lower the forgetting index only if you still have problems with recall and the materials is vitally important.

Why do we remember weird words?
(MRW, Mar 04, 2002)
Question:
Have you noticed that in learning English, the words that are easy to remember are often the "weird" ones (e.g. tintinnabulation)
Answer:
It all depends on the associations formed in your brain. Tintinnabulation may have stuck easily with you while it could be somebody else's pet peeve. However, weirdness often implies uniqueness which helps you avoid memory interference. Then, one day, you learn the word tinnitus and, all of a sudden, tintinnabulation may start causing serious trouble

"Successful" leeches should not be reset
(christian.roessel, Germany, Wednesday, January 10, 2001 11:26 AM)
Question:
What to with a leech that has been well-known for the last n repetitions? Is it more time-efficient to forget and reintroduce this element or shall I keep it as is?
Answer:
If the leech has been remembered for a number of repetitions, its inherent problem might have already been resolved. In general, you should take postpone, reformulate or delete actions only at the moment of forgetting the leech. As long as it progresses towards longer intervals, your best action might be to do nothing or to add minor edits that you believe could improve retention

You can learn Arabic with SuperMemo
Question:
Can I learn Arabic with SuperMemo?
Answer:
Yes. See: Arabic Verbs collection

Copying material from a dictionary (#17003)
(Rune, Norway, Monday, April 28, 2003 1:38 AM)
Question:
I copy word descriptions from the Collins Cobuilder dictionary and paste them into the answer field. It would be nice, if SuperMemo could create a new learning item and paste the description into the answer field. Now I first have to copy from Collins, create an new element, and paste into the answer field
Answer:
The best way to handle dictionary items is to paste the entire item to SuperMemo with Ctrl+Alt+N. Then extract individual definitions along with the headword with Alt+X. Finally, while learning individual definitions, create individual passive, active or detail items with Alt+Z

Here is an example of learning the meaning of the word trachea. Although there are 19 items on the output, not all these items are necessary to extract the basic meaning of the word. For that reason, the process can be executed incrementally. More specialized meaning can be refined in more advanced stages of learning.

  1. You start with the definition pasted from www.dictionary.com: 

tra�che�a   Audio pronunciation of "trachea" ( P )  Pronunciation Key  (trk-)
n. pl. tra�che�ae (-k-) or tra�che�as

  1. Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe.
  2. Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods.
  3. Botany. One of the tubular conductive vessels in the xylem of vascular plants.
  1. In SuperMemo, you clean up the definition to ensure only vital information is included:
trachea   
  1. Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe.
  2. Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods.
  3. Botany. One of the tubular conductive vessels in the xylem of vascular plants.
  1. With Alt+X you generate three extracts:
trachea   Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe

trachea   Zoology. One of the internal respiratory tubes of insects and some other terrestrial arthropods

trachea   Botany. One of the tubular conductive vessels in the xylem of vascular plants

  1. The first extract will be processed with cloze deletions as follows:

trachea Anatomy. A thin-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs. Also called windpipe

  1. This will result in the following cloze items:
a cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
trachea
trachea: A [thick/thin]-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
thin (thickness is a relative concept and you may want to skip that property)
trachea: a [bony/cartilaginous/muscle/membranous] tube descending from the larynx to the bronchi
cartilaginous
trachea: A cartilaginous tube [descending/ascending] from the larynx
descending
trachea: A tube descending from the [...] to the bronchi
larynx
trachea: A tube descending from the larynx to the [...]
bronchi/lungs
trachea: A tube descending from the larynx to the bronchi and carrying [...] to the lungs
air
trachea: a tube carrying air to [...]
(the) lungs/bronchi
trachea: A tube carrying air to the lungs. Also called [...]
windpipe
  1. The final list of questions and answers will look as follows:

Q: a cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
A: trachea

Q: trachea: A [thick/thin]-walled, cartilaginous tube descending from the larynx to the bronchi and carrying air to the lungs
A: thin

Q: trachea: a [bony/cartilaginous/muscle/membranous] tube descending from the larynx to the bronchi
A: cartilaginous

Q: trachea: A cartilaginous tube [descending/ascending] from the larynx
A: descending

Q: trachea: A tube descending from the[...] to the bronchi
A: larynx

Q: trachea: A tube descending from the larynx to the[...]
A: bronchi/lungs

Q: trachea: A tube descending from the larynx to the bronchi and carrying[...] to the lungs
A: air

Q: trachea: a tube carrying air to [...]
A: (the) lungs/bronchi

Q: trachea: A tube carrying air to the lungs. Also called [...]
A: windpipe

Q: zool: [...]: one of the internal respiratory tubes of insects and some other terrestrial arthropods
A: trachea

Q: zool: trachea: one of the internal[...](function) tubes of insects and some other terrestrial arthropods
A: respiratory

Q: zool: trachea: one of the internal respiratory tubes of[...](main animal group) and some other terrestrial arthropods
A: insects

Q: zool: trachea: one of the internal respiratory tubes of insects and some other [aquatic/terrestrial] arthropods
A: terrestrial

Q: zool: trachea: one of the internal respiratory tubes of insects and some other terrestrial [...](phylum)
A: arthropods

Q: bot: trachea: one of the[...] in the xylem of vascular plants
A: (tubular conductive) vessels

Q: bot: trachea: one of the tubular conductive vessels in the[...](tissue) of vascular plants
A: xylem

Q: bot: trachea: one of the tubular conductive vessels in the xylem of[...](division) plants
A: vascular

Q: trachea: one of the tubular conductive vessels in vascular [plants/animals]
A: plants

Q: bot: [...]: one of the tubular conductive vessels in the xylem of vascular plants
A: trachea

How can SuperMemo help an actor?
(David Baldwin, Saturday, August 17, 2002 4:45 AM)
Question:
As an actor specializing in one-man plays, I'm required to memorize and regurgitate hours of material. How can I best use SuperMemo to quickly and permanently memorize?
Answer:
SuperMemo is primarily useful in learning material that is to be retained in memory for months and years. It might be very useful in lifelong retention of quotes, citations or proverbs. However, if you need to memorize long passages, and you need to be able to recall them for weeks rather than years, SuperMemo's applicability will be limited. You could use it, for example, to cram pieces on which you stumble or get stuck. You could paste the problematic piece of text and generate lots of cloze deletions (i.e. questions with blanks to fill). With regular repetitions, the recall of such pieces of text will become quite automatic. You cannot expect much qualitative change in your performance as an actor by employing SuperMemo today. However, it is still recommended that you get to know it. Gradually, you will get a better grip on the inner workings of your memory. You will also find optimum ways of employing the program. With time, you might discover that SuperMemo is more useful that suggested in this answer. However, that will depend on your creative approach to selecting the right portions of knowledge to learn, and representing them optimally for your purposes

Strategic recall is a matter of knowledge structure
(Noel L, Sweden, Tue, 19 Mar 2002 13:11:25 -0000)
Question:
How can I bring knowledge learned with SuperMemo into my mind when I need it
Answer:
Ideally, the knowledge is already stored in your memory and it comes up naturally when you need it. However, the picture is more complex if you consider that the way you ask the question in SuperMemo may differ from the way your life asks you the same question. In other words, you may store some material in SuperMemo, but a real-life situation will trick you into being unable to recall it. In other words, you need to properly formulate the material to maximize its recall in all potential contexts (see: "20 rules" article). There are two basic tools that will help you keep knowledge at hand in need. The first is simplicity. Simple things are easier to remember, but they are also easier to apply in varying context. The second is universality of rules. For example, it is better to learn a universal mathematical formula than just the examples of its use. Examples can be used to emphasize applicability in various contexts. To sum it up: to make knowledge available at hand, you need lots of practical experience in using SuperMemo. See for yourself what works and what does not

Sequences are best learned through logical consequence
(Gary B., Canada, Mar 25, 2005, 19:37:03)
Question:
I am memorizing DNA synthesis. After formulating various extracts and cloze questions, how can I ensure that I am able to recall the events in the proper sequence? Here is the material I am learning:

DNA Replication:

  1. Once parental DNA is unwound and stabilized, replication fork forms at fixed origin of replication.
  2. One new DNA strand called leading strand is synthesized continuously as DNA polymerase moves toward replication fork.
  3. DNA is made in 5' -> 3' direction
  4. Because DNA polymerase only adds new nucleotide to 3', RNA primer (synthesized by RNA Polymerase) is needed to starts synthesis.
  5. Lagging strand of new DNA synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves away from replication fork.
  6. DNA polymerase digests RNA primer and replaces it with DNA.
  7. DNA ligase joins new discontinuous fragments of lagging strand

Answer:
All you need to do is to ensure that the sequence is a logical consequence of extracts you are processing. 

For example, the following sentence creates a link between events: Once DNA is unwound, a replication fork forms. The preceding event is unwinding DNA, and the succeeding event is the forming of the replication fork (naturally, you must know what a replication fork is before you go into details). If this exemplary sentence is well consolidated in your memory, you will have little problems with figuring out the correct logical sequence of events.

In other words, there is no need to memorize the sequence in a separate set of questions. If you understand the sequence, see the big picture, and consolidate individual steps, you shall have little problems with reproducing the events of DNA replication (today or 30 years from now). This guarantee requires primarily that: (1) you understand the text and (2) you generate some material redundancy to counteract the forgetting index. You generate redundancy with prolific cloze deletions. Here is how your material might look once incremental reading process is completed:

Q: Once parental DNA is [...], replication fork forms at fixed origin of replication
A: unwound

Q: Once parental [...] is unwound, replication fork forms at fixed origin of replication
A: DNA

Q: Once parental DNA is unwound, [...] fork forms at fixed origin of replication
A: replication

Q: Once parental DNA is unwound, replication [...] forms at fixed origin of replication
A: fork

Q: Once parental DNA is unwound, replication fork forms at fixed origin of [...]
A: replication

Q: Once parental DNA is unwound, replication fork forms at fixed [...] of replication
A: origin

Q: One new DNA strand called [...] is synthesized continuously as DNA polymerase moves toward replication fork
A: leading strand

Q: One new DNA strand called leading strand is synthesized [continuously / discontinuously] as DNA polymerase moves toward replication fork
A: continuously

Q: One new DNA strand called leading strand is synthesized continuously as DNA polymerase moves [towards / away from] replication fork
A: toward

Q: One new DNA strand called leading strand is synthesized continuously as DNA polymerase moves toward [...]
A: replication fork

Q: One new DNA strand called leading strand is synthesized continuously as [...](enzyme) moves toward replication fork
A: DNA polymerase

Q: Because DNA polymerase only adds new nucleotide to [3/5]', RNA primer (synthesized by RNA Polymerase) is needed to start synthesis
A: 3

Q: DNA is synthesized in [5'->3' / 3'->5'] direction
A: 5'->3'

Q: Because DNA polymerase only adds new nucleotide to 3',[...] primer is needed to start synthesis
A: RNA

Q: Because DNA polymerase only adds new nucleotide to 3', RNA [...] (synthesized by RNA Polymerase) is needed to start synthesis
A: primer

Q: Because DNA polymerase only adds new nucleotide to 3', RNA primer (synthesized by [...](enzyme)) is needed to start synthesis
A: RNA Polymerase

Q: Because DNA polymerase only adds new nucleotide to 3', RNA primer (synthesized by RNA Polymerase) is needed to [...]
A: start synthesis

Q: [Lagging / Leading] strand of new DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves away from replication fork
A: Lagging

Q: Lagging strand of DNA is synthesized in approx [...] piece nucleotides called Okazaki fragments as DNA polymerase moves away from replication fork
A: 1000

Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called [...] as DNA polymerase moves away from replication fork
A: Okazaki fragments

Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves [towards / away from] from replication fork
A: away

Q: Lagging strand of DNA is synthesized in approx 1000 piece nucleotides called Okazaki fragments as DNA polymerase moves away [...]
A: replication fork

Q: DNA polymerase digests RNA primer and replaces it with [...]
A: DNA

Q: [...](enzyme) joins new discontinuous fragments of lagging strand
A: DNA ligase

Q: DNA ligase joins new discontinuous fragments of[...] strand
A: lagging

Before you start learning, you would best ensure you understand the big picture of the entire replication process as your short sequence does not fully illustrate it. You can also build the big picture incrementally, but that may take substantial time and is not recommended when you learn for an exam. If you have the big picture in your mind before you begin to learn incrementally, the savings in time may be substantial. However, if you are learning from lecture notes, this rarely is a case (a testimony to the dismal efficiency of taking notes during lectures in the "old world" of learning). Luckily, you can easily substitute with plethora of well-written materials imported from the web.

For the HTML source of the incremental reading process that led to the above questions see: DNA Replication example

You can learn to retrieve words faster from memory
(Samson Chen, Aug 10, 2005, 01:17:00)
Question:
I learn English. I use items like this one: 

Questions: to move in a direction suddenly and quickly 
Answer: shoot 
Verb prep/adv 
eg. A car shot by. 

At the moment, I have no problem recalling the definition of a term. And if I see the word "shoot", I know one of its meanings is "move in a direction suddenly and quickly". However, I have trouble using these words in daily life. Although my oral English is fluent, I am unsatisfied with limited vocabulary in my speech
Answer:
The main problem with your item is that the question does not bring the word "shoot" instantly to your mind. A fluent speaker might answer your question differently (e.g. "rush", "dart", "hurry", "blast", etc. ). In other words, your question part is not specific enough. 

Your optimum learning strategy should include:

  1. a passive recognition of the word "shoot" (which you have already accomplished),
  2. the ability to use any term for "moving quickly" (e.g. the answer might then be "rush/dart/shoot" and any term would qualify for a good grade), and, optionally,
  3. some enriching items that would help you use various synonyms and enliven your language.

For the latter, you could use cloze deletion examples that would anchor a specific synonym in a specific context. For example: "The police car ... by in a chase", or "Johnny English ... out from a hot tub". 

Passive items will help you understand other people. Active synonyms will help you express yourself. Anchored synonyms will help you enrich your language. 

You may memorize a couple of cloze examples for the word "shoot". It is important though, that you do not rigidly cram them if you fail to anchor "shoot" in a given context. If you fail a few times, change your example. Probably it is not specific enough.

Some narratives may be thin on memorizable material
(Tanya, Sunday, November 24, 2002 11:46 PM)
Question:
I am trying to learn incremental reading. For starters, I decided to learn the incremental reading FAQ using incremental reading. However, I have no idea how I might cope with FAQs such as " High priority of material or long review intervals will prompt you to run an article preview". The answer is too long and I don't know how to grab it.
Answer:
Not all materials are suitable for generating extracts and clozes. Some narratives should just best be read passively. Like the quoted FAQ, they may be a compilation of facts that are generally obvious. In such cases you can just read and dismiss. Or you can read and schedule another review in a month or in a year (if you worry you miss something important). Or you can try to write, in your own words, a sentence or two on what new things you have learned from the narrative. Your sentence would shortly extract the quintessence from an otherwise lengthy passage. If it is meaningful and quintessential, you shall find little trouble with locating keywords suitable for clozing. 

Nevertheless, if you really wanted to process that particular FAQ with incremental reading, you should eliminate as much verbiage as possible, avoid memorizing it as an enumeration, and focus on its components that are the least obvious. In such a case you might arrive at the following set of questions and answers:

Q: The most important incentives for whole-article preview: [low/high] priority of the material
A: high

Q: The most important incentives for whole-article preview: [short/long] inter-review interval
A: long

Q: The most important incentives for whole-article preview: [presence/absence] of higher-priority fragments buried in a lower-priority text
A: presence

Q: Extract-preview will [increase / decrease / not change] your exposure to previewed material
A: increase

Q: Line-at-a-time reading will be equivalent to assigning a [lower/higher/same] priority to an article (assuming you do not interfere with intervals, A-Factors, priority settings, etc.)
A: lower

Q: If your reviews occur in very long intervals as a result of slow reading, you may opt for [...](passive method) or running a preview of the most important sections instead
A: shortening the interval

Q: If your reviews occur in very long intervals as a result of slow reading, you may opt for shortening the interval or [...](active method)
A: running a preview (of the most important sections)

Q: if you are reading texts from your e-mail tasklist, preview is highly recommended because [...]
A: not everyone starts their message with the most important points

Inventiveness is the key to mnemonic techniques
(Daria, Jan 12, 2006, 19:13:30)
Question:
I could never remember the name of tryptophan. I remember its name just more or less. Do you think it is a good idea to add an item like this: 

Q: [Hit the fan] is an amino acid that is essential in human nutrition 
A: tryptophan

Answer:
Yes. This item will work, esp. if "hit the fan" is what comes to your mind when speaking of tryptophan. However, your item does not say anything characteristic of tryptophan itself (many other amino acids are essential as well). In other words, you are only learning to associate the name and "hit the fan" sound. You could fare much better in the long run if you used the formula of tryptophan as the question (e.g. "What amino acid is presented in the picture?"), or better yet ask a question based on something unique or memorable about tryptophan. For example, you could ask "What amino acid used as a supplement caused an autoimmune disease outbreak due to supplement contamination (US, 1989)?"

Math proofs in SuperMemo
(JOSEPH PRIMO BELARMINO, Feb 06, 2006, 03:18:38)
Question:

I'm trying to format math proofs into SuperMemo, but I'm not sure exactly how to make it so that it's always shown sequentially.
Hence if I have a proof that goes like:

often I'd get the step 2 window such that I'm unintentionally getting information for the step 1. How do I fix this?

Answer:

You can decompose the proof into individual cloze deletions. For example:

Q: Proving X:

The sequence of review is not important. Repetitions are not supposed to help you build the big picture of the proof. They are just supposed to refresh memories on the assumption you have already built the coherent picture of the proof. The more clozes you create, the less likely you are to lose the coherence of the proof. Even for simple proofs as above, it may often be necessary to generate 6-8 separate cloze deletions. Moreover, individual steps are usually a bit more complex and need to be clozed too.

The difficulty is only in making sure that your clozes are well-structured and do not provide hints that make answers obvious. For example, you do not need to worry about the hint given by Step 1. After all, in real life, you will always worry about Step 2 only once you solved the Step 1. However, it may be necessary to delete Step 3 from the question as it may hint on the action taken in Step 2. You can, for example, move Step 3 from the question to the answer in the form of the comment:

Q: Proving X:

A: b=c (step 3: using theorem d,  c is the final answer)

It will also be helpful if you try the entire proof in a real-life situation to make sure your clozes indeed do the job of making you able to prove the theorem.

See also: Sequential review may be inefficient

Syllables may be easier to remember than numbers
(stephanie phillips, Jan 14, 2006, 21:40:55)
Question:
Is it easier to remember a sequence of letters (example aw by hu il te be ha qx lm ni) rather than remembering a sequence of numbers (example 33 57 96 75 29 65 92 01 75)? If so why?
Answer:
No theory answers such questions better than a practical test. If you put a couple of examples of both sequences to SuperMemo, you will be able to compare performance in a number of ways (for example, an average inter-repetition interval after a longer period of time). Such tests are important as individuals may differ substantially in the way they convert question-answer pairs into specific representations in their mind. Moreover, there is substantial interference between student's knowledge, knowledge currently learned and retained with SuperMemo, and the newly memorized material. If syllables are easier to recall than numbers then, as in all similar cases, it is related to the complexity and uniqueness of individual memories that are to be consolidated. Numbers are less unique and an average individual does not store specific representations beyond the ten digits, plus a set of numbers (s)he easily associates with other memory landmarks such as one's own birthdate, the date of landing on the moon, etc. Syllables, on the other hand, are central for language processing and a typical set used by an average individual (including "sounds-like" syllables) will be far larger than the set of landmark numbers. The ability to store memories in a specific form is related to their evolutionary applicability. Humans have used basic numbers in a limited way for a couple of millennia. That time-span is too short to equip our brain with specific number circuits. Spoken language has been in use for a period 10-40 times longer. And the best mnemonic tool, visualization, capitalizes on circuits that have been in use for hundreds of millions of years (i.e. even before the appearance of vertebrates). The importance of uniqueness of memories is well illustrated by the fact that you can best defeat a mnemonist by asking him or her to memorize strongly interfering patterns. For example, when asking to memorize objects, you can list those that are very similar to each other, e.g. rod, javelin, stick, pole, whip, pale, spear, pike, etc.

Language ambiguities can be tackled with SuperMemo
(G.W., Mar 09, 2007, 03:26:54)
Question:
I learn Spanish. Spanish words often have multiple English meanings. When all common English meanings are matched to the same Spanish word, failure rate dramatically increases. When single meanings are matched, much is lost
Answer:
When multiple meanings slow your progress, you could define Spanish words using Spanish definitions. Thus you can ask for two different meaning of the word "album" with two distinct questions:

This approach should effectively cover for ambiguities in active recall. As for passive recognition, you can use synonyms with appropriate context. For example:

As your fluency increases, you can hone your semantic sensitivities with exemplary sentences processed with incremental reading. Here is another meaning of "album" taken from a real life situation:

We sat with my husband browsing our photo [...](book with pictures) from Madagascar.

Pick those sentences from real life situations in which your original active or passive items did not do their job up to your expectations.

If sequence is important, it must be included in a single item (#348)
(Martin Zielinski, Friday, January 21, 2005 12:27 AM)
Question:
How can you memorize a poem if SuperMemo keeps randomizing the entries I input
Answer:
If you need to recite the entire sequence in one go, you must simply include the entire sequence in the item's answer. 

It is economical to use SuperMemo for poems only if you need to be able to recall them for long periods of time. If you learning process is sparse (few repetitions, few postpones), you might consider using SuperMemo for poems if you need to retain them for months. However, a heavily loaded learning process may extend that minimum period to years. Learning poems is simply "workload intensive" and will stifle progress in all other areas. 

As for the extraction of memory bottlenecks, the optimum strategy will depend on your goals, available time, and the required fluency. The higher the fluency required, the greater the redundancy needed. What makes learning poems costly is the fact that redundancy grows exponentially with fluency. For high levels of fluency, traditional poem learning techniques may be necessary and frequent recitation unavoidable. 

If you want to apply SuperMemo, one item must include the call to reproduce the entire poem and must use a forgetting index that will determine the probability of recalling the entire poem. Without this item, you will always be able to reproduce fragments of the poem, but there is a possibility of the disruption in the chain. That cumulative probability will be very high for a large number of paragraphs (each carrying its own probability of forgetting). 

In addition to the main poem item, you should include individual items in which the preceding line evokes the succeeding line. Those items will assist automation of the recitation and plug up weak-link bottlenecks. The hardest fragments thus will be repeated more often. 

For example: 

Q: Look, how spry she still is, how well she holds up: 
A: hatred, in our century. 

Q: Look, how spry she still is, how well she holds up: hatred, in our century. 
A: How lithely she takes high hurdles. 

Q: hatred, in our century. How lithely she takes high hurdles. 
A: How easy for her to pounce, to seize and so on. 

If it appears that one of the items above still causes frequent problems (e.g. wrong recall, substitutions, wrong word order, etc.), you will need to generate more specific queries. 

For example: 

Q: Look, how [...] she still is, how well she holds up: hatred, in our century. 
A: spry 

Q: Look, how spry she [...], how well she holds up: hatred, in our century. 
A: still is 

Q: Look, how spry she still is, [...]: hatred, in our century. 
A: how well she holds up 

The fastest way to generate individual items is to use incremental reading tools. In addition, if you are not in a hurry, the entire process should be incremental. This means that you should generate detailed items only from fragments of the poems that you frequently stumble on. Processing the entire poem at once will generate an astronomical number of items and be very costly. Such gained redundancy will improve recall, but may make the entire process uneconomic

Vocabulary items for SAT exam (#11208)
(�� �3��, Nov 07, 2006, 15:00:34)
Question:
I'm using SuperMemo to memorize college-level vocabulary (Barren's 3500 basic word list (for SAT)) which includes the hand-made
flashcards for long-term memory. I'm now memorizing 40 words per day (1000 words in a month). Can you comment on some of my items?

For example:

Q: badger
A: pester [Korean translation of pester]; annoy
"Her boss badgered her to do the sales report early."

Answer:
As this is your first month, it will take you yet some time to discover your own mistakes. Although your item will work, it is far from ideal. Here are a few suggestions:

  1. ambiguity: if you learn the verb badger, precede it with to to avoid confusion with badger the animal
  2. conglomeration: you add lots of information into the answer. This will make learning a bit harder. You would better create separate items for synonyms, Korean translations and examples

If you produce several simpler items, you may actually spend less time on learning and yet produce memories that might appear more useful at the exam.

The simplest way to proceed is to use incremental reading. You could start from writing down the essential information in a single topic:

to badger = pester/annoy/nag (e.g. "Her boss badgered her to do the sales report early.")
pester = [Korean translation of pester]

Using extracts and cloze deletions you will only need a few clicks to produce three items:

  1. Passive: 
    Q: to badger = [...]
    A: pester/annoy/nag
    This item will help you recognize the verb badger. It is not important if you answer pester, annoy or nag. Any synonym will qualify as the correct answer
  2. Active:  
    Q: to [...] = pester/annoy/nag (e.g. "Her boss [...] her to do the sales report early.")
    A: badger
    With the help of the above example, you can try to actively use the word badger. This may not be essential for SAT, but will definitely improve the quality of your English. Note, however, that there are many other words that would fit this context. If you keep answering differently, you will either need to change your item or allow of several synonyms for this particular context
  3. Korean: 
    Q: pester = [...](Korean)
    A: [Korean translation of pester]
    You should not insert Korean into your all-English items. If you see the word pester in the answer, and you fail to remember its meaning, you should search for the occurrences of pester in your collection and do a subset review to refresh the word and grade yourself Fail or less on the third item (and others, if they cause trouble due to the weakening of your memory of pester)

Try to mimic real life situations to combat memory interference
(Dave M., Sunday, April 29, 2007 12:59)
Question:
I would like to ask your advice on how to deal with memory interference. I’m a 3rd year Osteopathic medical student. Unfortunately, this already means I have to contravene some of the rules of learning that you’ve listed. I’ve been using SuperMemo for the past couple of years now but I’m having a particular problem with some of the information that has to be remembered for ‘viva voce’ exams. An example might be the sympathetic nerve innervations of the visceral organs. From a diagnostic point of view I need to know that, say, someone with a liver problem may have a spinal dysfunction at the level of Thoracic vertebrae level T7-9. Or alternatively, if I see a dysfunction at this level, that there may be a liver dysfunction in the patient. Sadly, there are many similar overlapping items in this sort of category leading to significant interference. 
Below are some examples of the Q&A pairs that I’ve put together so far.

Example 1:

Q: Which segmental level is the sympathetic innervation of the heart?
A: T1-T5
Q: Which segmental level is the sympathetic innervation of the bronchi and lung?
A: T2-T4
Q: Which segmental level is the sympathetic innervation of the oesophagus (caudal part)?
A: T5-6

Example 2:

Q: Which organs receive sympathetic innervation from the L1-L2 segmental levels?
A: Splenic flexure to rectum
Q: Which organs receive sympathetic innervation from the T10–L2 segmental levels?
A: Lower limb
Q: Which organs receive sympathetic innervation from the T6-T10 segmental levels?
A: Stomach, spleen and pancreas

Answer:
All your items seem to carry important information, and yet they are truly scary due to their similarity. If you generate more than 10 items per each example, memory interference is likely to be astronomical. Such items will quickly join the pool of "leeches" (items that are very hard to remember), and will clog your learning process slowing you down in all other areas. In the end, for your future as a medical student, it might be better to never include these in SuperMemo (as if you never had access to SuperMemo in the first place). 

Luckily, there are proven techniques that will help you tackle similar knowledge with SuperMemo. All solutions are costly, but will pay handsomely in the long run. The basic principle is to gradually glue individual pieces to your overall knowledge structure and be as visual and mnemonic as possible:

Uniqueness of memories helps retention (#3855)
(Mike Condron, Wednesday, November 29, 2000 1:29 AM)
Question:
Some esoteric facts I can hold with ease (e.g. 2 molecules of tetrahydrofuran are used in purine synthesis) but others I can't (e.g. which step in the TCA cycle makes FADH?)
Answer:
If you make unique associations, they are subject to far less interference. As such, they are easier to remember. In the presented example, you may form a unique association: tetrahydrofuran-purine. However, you may recall many pathways in which FAD plays a role. Even if you remember which reactions are catalyzed by dehydrogenases, you may wonder if FAD or NAD plays the role. In such hard cases, you may try to use mnemonic techniques. Try to imagine FAD as your pet TCA torturer, and SUCCINATE as your smiling face expressing SUCCESS. With this link, you may need only 5-6 repetitions to retain FAD-succinate pair for lifetime

Sequential review may be inefficient
(name zero, Sep 23, 2010, 20:11:59)
Question:
When doing repetitions, there are sometimes two elements that must be presented to me together, even in a particular order, e.g. questions like Element 1: "What do you need to create a beam?" : "SomethingX", and then the other element, Element 2: "What are the properties of SomethingX?" : "some properties as answer". If just the second question was presented to me it might be so much out of context, or there might even be ambiguity, that I cannot answer the question properly. I searched the help and there you recommend to cure problem b) by creating bigger elements with more questions/answers in one element. I don't really like this solution I'm afraid. There doesn't seem to be anything in SuperMemo do achieve this, so that's why I am suggesting it as feature request
Answer:
If you want to semantically connect a group of elements related to a single subject in incremental reading, you can use subset review based on the elements tree structure (see below). This way you can quickly review all elements related to a topic of which, you believe, you lose the grasp on the big picture. However, you should remember that you will always minimize the review time, if you ensure all elements are self-contained with their full context, and are reviewed independently. Each review of a knowledge granule belonging to a larger well-connected structure entails benefits to long-term retention, as well as benefits to selected components of the structure itself. In your case, if you have SomethingX well defined and remembered (in the ideal case), you should not have problems with listing its properties. Note also that a list of properties, is obviously a list (or a set), and should be avoided in the learning process if possible due to its high expense. Ideally, you would create topic with properties of SomethingX and generate a number of cloze deletions that would help you recall all individual properties independently.

As for the procedure for your desired sematic review, you can, for example, do the following:

Note that abusing the above procedure will add extra time to your long-term learning process, and increases the chances you will forget individual components of the structure due to the spacing effect that comes into force with premature review. A slightly more efficient variant is to review only a portion of the subset in question. As long as the big picture re-emgerges in your mind, it is advisable to move on with your normal (i.e. not structured) learning process. When reviewing the portion, depending on circumstances, you should sort the subset along the knowledge tree (if sequences are important), along the priorities, or randomly (for large sets, in which high priority elements are well consolidated and should not be given preference).

You are best at judging your own item formulation
(M van der Laan, Netherlands, Nov 04, 2010, 00:33:25)
Question:
In a lot of cases, I don�t ask the question in a question-like format. Instead I may use a statement:

Q: syn: supercomputer sound
A: lfo=pitch > lfo s&h

I know when I see this particulair item I need to answer how to make a supercomputer sound. But it could be any number of things, as the statement is not specific to what it is referring to.  
1) does it influence memorizing and applicability when you use a statement instead (of a question) in the question-field?
2) does it influence memorizing and applicability when the statement is not completely specific as in the example above?

Answer:
If this method works for you, there is no reason to worry. Your brain clearly does not have any problems with converting your statement into a meaningful question. Short questions might even speed up the processing of items. If you ever start having problems with recognizing the question in the distant future, you can always replace it with a longer sentence.

Some texts are too general to be handled efficiently with cloze deletions
(Carl S., Sat, 13 Apr 2013 16:10:12 +0200)
Question:
I did read all the 20 rules for knowledge formulation, and I think I understand them, but I am not sure how to properly apply them.

Example of piece of knowledge: A good enterprise architect should enable the right balance between the needs of the organization for an integrated IT strategy, permitting the closest possible synergy across the extended enterprise, and allowing individual business units to innovate safely in their pursuit of competitive advantage.

I have no problem understanding this phrase (synergy across business units must be balanced against freedom to innovate within business units), and I constructed two cloze deletions from it for either side of the balance, and when presented with either question I can fill in the blanks. Yet, I was in a discussion recently defending a synergy position, not realizing that it might jeopardize innovation, not even realizing I had this SuperMemo question pointing it out to me.

In other words, even though I can answer the question in SuperMemo, it is not something that stuck in my memory, i.e. the synergy is not associated with freedom-to-innovate and vice versa (maybe it will after this mail). So, I suspect that I am only able to answer the question based on recognition of the question or some such, and not on recognition of the association.

I figure that a better way to associate the two would be to ask something like "what must be balanced against each other" but this question would be so general in nature that it would create serious interference with other questions that deal with other aspects of enterprise architecture that need to be balanced against each other. Or I would have to make it more specific, again risking to give away the answer in the question, which would also not cause the association to form.

This question is but one example, but I have lots of such cases of hidden (as here) or explicit enumerations where I need to consider many elements together, and I fail to formulate my questions in such a way that these elements come together in my mind when I think about the subject.

Answer:
Your conclusions are correct. In similar cases, you need to pause to ponder what kind of questions you want to be able to answer having read the passage. If the questions are too general or too obvious, you just need to trust your own intelligence and creativity to be able to answer them on the basis of your experience and more specific questions in the given area of knowledge.

In this case, keywords such as "synergy" or "innovate" might provide a hazy way to capture the meaning of the passage. However, very general texts are not suitable for treatment with cloze deletions. You may waste unnecessary time on re-reading the entire passage at questions time, or waste time on simplifying the passage to capture the essence. In school jargon, you might call similar passages "waffle". They may carry an important message, they may help the flow of text, they might be explanatory, but they do not yield material suitable for memorization. In the extreme case you can juxtapose Wikipedia-like "IT = Information Technology" with "waffle" that cannot be clozed: We should be nice for other people.

In all sorts of exams, you will always need to tackle lots of "waffle". You will also meet teachers who demand fluent "waffle" performance. However, this is not the type of knowledge that will make you a better expert or a better person. If you meet "waffle", pause to think if there are questions that truly flow from the text, or if the text is too general to be handled with SuperMemo. In your case, you might do better by perhaps adding some meatier passages on enterprise synergy or constraints on innovation or ... Actually, you are the best person to find supplementary material that will help you better understand the underlying issues.

If you are new to incremental reading, working with texts from Wikipedia can be an excellent training ground. Due to the nature of crowdsourcing, Wikipedia lends itself perfectly to incremental processing. Once you get the hang and feel the benefit, you will quickly learn to spot text and passages that are less suitable and provide less benefit when processed with incremental reading.