Collaborative Learning.page

# Questions and ideas

* How to get questions to the right people?
* How to leverage crowdsourcing? -- difficulty, categorization, dependencies, component tagging
* Can we tie in question-answering with reviewing?
* How to encourage participation? (Examples: [LiveMocha](Online Language Learning Communities), StackOverflow)
* Can we use ideas from markets or algorithmic game theory?

# Theoretical basis for collaborative learning

This exercise is to explore collaborative learning from the ground up, attempting not to bias towards using Learnstream in any way. Given that it may be possible to learn by doing, we'll start with collaboration in general.

Generally speaking, some kind of task will have inputs, the things available on the onset, and outputs or goals, the things that the doers of the task intend to accomplish. 

We'll look at two inputs:

* I1: structured education, where the structure is predetermined by something like an instructor giving lectures or a textbook
* I2: unstructured collection of info

and two goals:

* G1: each individual learns the content and ideas of a particular subject
* G2: completing a task 

Examples:

* G1: playing an instrument, G2: collaborative music making
* G1: learning a programming language, G2: collaborating on a code project
* G1: learning physics, G2: advancing the field of physics through research

Note that G2 is basically a generalization of G1: We can define the task to be that every student learns the subject equally well. The advantages of the G2 examples above are that things useful to society are being produced, people are learning , they might be more motivated etc. One disadvantage is that someone can get out of the loop and not be able to reach full potential unless extra care is taken to educate them. 

Examples of G2 style learning: medical residence; Harvard's Math 55, where students essentially write a textbook; job training, oftentimes

# A general method for collaborative learning

Let's look at I2G1 for now. 

1. Bits of unstructured info are distributed among the students
2. Highlight unknown ideas, add questions
3. Rate comfort w/ terms and concepts (begins some kind of SRS thing to instill these to memory)
4. Use graphical model to determine hierarchy of prerequisites (probabilistic) (either auto or manual) E.g. to understand recursion you must first understand recursion
5. Generate examples and exercises

Collaboration Inspirations: science, wikipedia, open source software