Minimum definition of incremental reading

From SuperMemopedia
Jump to navigation Jump to search

Why need a definition?

The term "incremental reading" was born in 2000, and it now slowly percolates to the awareness of users of spaced repetition. To ensure that no half-baked product can claim the title, please demand that the authors of software that claims to "implement incremental reading" comply with the "minimum definition" outlined below

Definition

First Steps options

  • spaced repetition
  • extracts
  • cloze deletions
  • read points (bookmarks)
  • priority queue (missing from SuperMemo 2004)
  • repetition auto-sort and auto-postpone (missing from SuperMemo 2004)
  • extract and cloze hierarchy in the knowledge tree (necessary for semantic review)
  • easy imports from the web (at least HTML and images)
  • HTML rendition (WYSIWYG)
  • propagating images

Vital Options

  • mid-interval review with a correction for spacing effect (this is solved in Algorithm SM-17)
  • early review tools (e.g. Advance, Add to Oustanding, etc.)(missing from SuperMemo 2004)
  • semantic review tools (esp. search&review, branch review, subset, etc.)
  • overload tools (e.g. Mercy, Postpone, or similar)
  • propagating references
  • image download/localization

Other options available in SuperMemo

  • A-Factor-based article review algorithm (or similar)
  • processing attributes and formatting (e.g. extracts, cloze keywords, highlights, ignored texts, etc.)
  • source-linking and hierarchy
  • visual learning (image zooming, cropping, compressing)
  • wholesale file&folder imports (with folder hierarchies)
  • web search tools (e.g. for easy imports from dictionaries and favorite data sites)
  • popular site filters (Wikipedia, YouTube, etc.)
  • tools for automatic decomposition of data (e.g. article splitting, dismember, decompose, etc.)
  • more formats for incremental learning: video, audio, mail, etc. Also: PDF, PowerPoint, and OneNote, although it is hard to demand this as part of the definition, esp. that SuperMemo fails to support these
  • progress statistics

Comments and Questions

SuperMemo 2004 is incomplete

mh wrote: By the above definition, SuperMemo 2004 does not live up to your promise "incremental reading freeware"!!!

SuperMemo 2004 is still solid

Without the priority queue, SM2004 is indeed lame, but it is still a good introduction to incremental reading. Once the volume of knowledge starts increasing, the user can upgrade to newer versions with the priority queue (those will become freeware at some point too). Necessity is the mother of invention. Early versions of incremental reading were not suitable for high volume learning and the importance of priority queue is hard to appreciate before actually using it. This is why it is absent from SuperMemo 2000 through SuperMemo 2004, while deemed indispensable for incremental reading by today's standards.

Update 2016

As of 2016, SuperMemo 15 is also freeware and it fully meets the criteria.

User experience is vital

mj wrote: Incremental reading was born in SuperMemo. I believe that nobody should claim to have implemented incremental reading without using it first in SuperMemo himself. The theoretical description is very different from practical applications. I believe that the text about IR is very short, tools used in IR are much larger, and the practical training for doing it well is even larger. This means that IR is two orders of magnitude bigger than that was is written about it at supermemo website. I cannot imagine implementing IR without using it for 2-3 years first!!! It is like designing a new sports car without ever sitting behind the wheel (of any car)!

How does topic algorithm work?

See: ABC of incremental reading for any user of spaced repetition