Open-Source and Dynamic Multi-Language Learning System

[navaburo]

I propose creating the ultimate self-study tool for beginner and intermediate learners of any language. I post about it here because (1) Putonghua will be a critical language to this project and (2) I respect the ideas of people on this forum and value your feedback.

This project seeks to provide a course which will take an absolute beginner and bring them up to a low intermediate level, and then assist through the intermediate phase. For the moment I will assume that the learner is already familiar with the script and the sounds of the language. What remains is to provide a package of core content covering a broad range of everyday topics and including implicitly the first few thousand most frequent vocabulary items and the most useful grammatical patterns.

In terms of high-level desired features, the project will be:

1. Free: licensed under Creative Commons Share-Alike Non-Commercial or similar
   
2. Structured: high-level flow of language level and topics is defined separately from the detailed content
   
3. Crowd-Sourced: structure and content provided by the community
   
4. Accurate: a moderation system (also crowd-sourced) will ensure accuracy
   
5. Multi-Language: many languages can be supported easily (it's just one more column in the database)
   
6. Dynamic: as new content is approved, it becomes available for learners to use
   
7. Multimedia: pictures, text, and audio for each item
   
8. SRS-based: reviews are scheduled so as to optimize memory storage strength
   
9. Unambiguous: pictures and translation disambiguate meaning. No vocabulary taught outside meaningful utterances (with possible exception of concrete nouns).
   
10. Portable: usable from traditional PC platforms and mobile devices
    

Right now the biggest implementation obstacle I have is related to #2 above: How can a logical and pedagogically-sound structure be maintained when the content is crowd-sourced? And, to what extent can the structure also be crowd-sourced? This definitely needs some theoretical work. I especially welcome suggestions here. When this structure concept is solid, all the rest will fall into place.

For the content contribution system, I imagine a web interface would be suitable. Perhaps a custom backend, perhaps a wiki engine? It is important to make content contribution accessible and fast.

There is also the question of how this will be deployed. It would be easy to maintain a database that can export to Anki decks, but it will be important to show content in a logical order (i.e., following the structure). Perhaps tags can be used to do reviews on certain topics or units of the course, so that the learner is not blasted with a grab-bag of unrelated reviews every time they study. The problem with using Anki is that we don't have high-level control over the topic flow... Major brainstorming is needed here too.

---

Thanks in advance for your opinions, suggestions, criticisms, encouragement, and other assistance,

- Chris

[imron]

IMO, if you rely solely on crowdsourcing from the get-go you will have little chance of success. See for example here.

That's not to say don't do it, but rather that like most successful Open Source projects it needs a committed person with a vision who will produce enough of an initial bulk to get the ball rolling and then keep pushing the project forward regardless of whether other people are interested or not.

For the issue you raised about item #2, my suggestion would be to come up with the pedagogical structure that you want first, and not let the design of this be crowdsourced (at least not initially). Have this as the overarching vision of what you are trying to achieve and what content you are looking to have, and how people should work through that content to make the most of the resource you are providing. Once you've got that, then you can have an overview page with %complete markers for each section so people know what stuff needs work/help and then it's just a matter of saying ok, crowd, go out and make the content we don't have yet.

Don't forget though that if you want the project to maintain momentum someone needs to be there plugging away at filling that content even if there's little crowd participation in order to avoid the chicken and egg problem of not getting much much crowd help if the project doesn't have any useful content, and not getting much useful content if you don't have any crowd help.

You can then have the ability for people to add content outside the main framework if they want (in a separate area of the site perhaps), as well as the ability for people discuss/vote for things that should be part of the main pedagogy and so on.

[gato]

This sounds like an online textbook for beginner students of Chinese at its core. Who do you see as the main contributors? I would think that the ideal contributors would be experienced teachers of Chinese.

What makes you think a wiki approach to this would produce better results than a team of professional teachers?

[Scandinavian]

Great initiative.

In terns of delivery, in my opinion the need for mobile outshines the need for desktop. My main constraint for learning Chinese is time, however, when I am on a bus or other place with limited internet connectivity, I usually also have time.

I am sure you have looked at what is already out there of free resources. There are several app available that will give the same amount of standard sentences to learn from, numbers, basic restaurant phrases etc. And once through these nothing more is offered. These basics should probably always be covered in any self study program, but there is need for taking it further.

Crowd Sourcing the content will hopefully get a good portion of diversity.

[navaburo]

Imron:

You are definitely right about the weaknesses of crowd-sourcing. I plan to develop structure and content myself for English/Chinese first, a kind of template, and then elicit help for recording audio, quality control, and and expanding to other languages.

I would also like to have the system flexible enough to allow for multiple people (say a small group of experienced Chinese teachers) to extend that template.

It is definitely desirable to have an offline mobile version. Some advanced features like speech recognition may require Internet though.

This is obviously a huge undertaking, but I've been thinking about it for a few years now and I have years more (hopefully) to put it all together. I just need to plan it out well up front.

Thanks for the suggestions

[renzhe]

Like I wrote in the thread imron linked to, devising a good language programme is a hard thing, and simply having learned a language does not automatically qualify anyone to teach. If you are going to develop structure and contents, I hope that you are qualified enough to do this. The success of your idea will hinge on this.

On a very relevant related note, have a look at Chinese Grammar Wiki. It's Creative Commons, and dedigned by professionals. At least it would make sense to join forces.