>With a bit of thought it may be possible to make a framework into which volunteers add a lot of the content, though editors/quality control would likely be necessary.
I like the idea of students up or downvoting stuff based on how clear it is. Even a diagram or a one-sentence description could have up and down votes associated with it.
We would have to take two factors into account, I think, those are; (1) clarity (2) accuracy. An “entry” can be very easy to understand but completely inaccurate, or completely accurate and horribly difficult to understand, and so forth. So one could vote on clarity and/or accuracy, depending on whether you are competent or incompetent.
I envisage going onto this site and thinking “I want to learn something new about the Fibonacci sequence”, or: “how does Quicksort work?” and so: “show me results tagged ‘quicksort’”.
One could shrug this idea off as a “Wikipedia clone-- with voting”, but Wikipedia is just an encyclopaedia; it does not aim to teach you things. But, we can take some ideas from Wikipedia...
If entries -- which would essentially be factoids, as I am hypothesising -- contain a few paragraphs and cover one very narrow topic, that would necessitate cross-linking of entries. Just as you read Wikipedia and follow hundreds of links because you just “have to!” know more about this topic, this could be applied to entries. Perhaps even “related entries” or “series” (an ordered set of related entries) could be created.
This is a very exciting idea to me. Does anyone know of anything like it? I may implement it if I see nothing like it. Let me condense the criteria:
1) A web site for learning things, however popular or esoteric
2) Learning is done by reading entries, submitted anonymously or otherwise.
3) Entries are short, to the point, and very narrow. (“What is polymorphism?”, “How do I implement RAII in C++?”)
4) Entries are voted upon by two factors indicating overall quality; (1) clarity (2) accuray.
5) Cross-linking, like a wiki.
Anyway, I got very excited about a very similar set of ideas. I'm not much for programming (yet) though. But go for it!
Here is an idea for dealing with prerequisite knowledge:
Let's say I want to learn about quicksort. For quicksort, let's say I need to know about arrays and recursion. For recursion, I need to know about functions.
If I search for quicksort, the program should analyze my user profile to see which prerequisites I need. Then it should generate a page with all the prerequisites and quicksort.
Now let's say I'm reading about recursion. At the end of the page it should ask me a few questions about recursion, for which I will type in the answers. Getting the questions correct should be a very good indication that I understand the section. If I get them wrong, the program should try to give me a hint. If this has not been implemented for a particular section, I should just get a generic extended explanation. Either way, I should have another question to answer.
Thinking about this is making me become mildly excited about my idea again. Maybe I will beef up my programming skillz and attempt an implementation.
Anyway, I emailed the author of <a href="http://eloquentjavascript.net/">Eloquent Javascript</a>, and here's what he had to say about computer-assisted learning:
>There is a lot of great
stuff that can be done with 'hyper-text', but it is tricky. My initial
plans for Eloquent JavaScript were way more extravagant, but going
away from the old style of text turns out to be rather hard -- that
form has been evolving for three thousand years, and is much enriched
by the fact that we are all used to it and know its conventions. When
a text is no longer continuous, you can not refer back, which somehow
makes it a lot less personal -- I liked that fact that, in writing
chapter 4, I could keep talking about stuff mentioned in chapter 2. A
radically 'active' text is bound to be either a disconnedted cloud of
snippets, or some carefully crafted non-linear experience that would
take years to get right.
I like the idea of students up or downvoting stuff based on how clear it is. Even a diagram or a one-sentence description could have up and down votes associated with it.