Friday, August 22, 2008

Predictive Learning Systems For Social Networking Portals

The social networking sites today holds huge amount of members. Based on my other post in e-Learning , I would like to suggest a new idea for Social Networking Portals too.

The social networking sites can add a new functionality, say "Way Of Life". So every user will have their own way of life and their own interests. This information can be shared among users after proper authorization from all the users.

When a user logs in, the "My Way Of Life" option will appear and they can share their interests, hobbies, likes & dislikes, Passion etc., So users can find people with similar interests and define their own way of life. This may be implemented in some portals with minimal features, but with a proper adaptive algorithm can identify quality living paths, which will lead to the betterment of human way of life.

I have serious doubts on this idea since people may misuse it to the core.

If you find any mislead opinion in this please share here.

Thursday, August 21, 2008

Learning Made Easy

360- e-Learning

e-Learning is nothing but easy learning.
I would like to have a small discussion on easy learning. This may not be new to everybody but I just want to publish this.

The whole idea starts with the term "Simplified Learning Patterns". When I started learning about e-Learning, I found hard to learn things. I couldn't find a simple path to learn things, since the content is massive. I couldn't find a definite way or path to take to accomplish learning about e-Learning.

This is an idea which I felt will make learning life easier. I haven't checked whether this idea exists in this world of web.

Simplified Learning pattern

This idea is applicable to a portal which is as huge as wikipedia, yahoo Answers etc

Here We go,

Assume there are some 100,000 users in a portal where the knowledge base index is over the regular limit. Huge amount of data.

Pre - Requisites
  • The content in the portal should have a metrics to measure its quality.
  • All the registered users should have a user rating based on their content quality, participation in various discussions & Quality of response etc
  • The portal should have a well defined terms & conditions approved by all the users for recording their learning history. Its more similar to web history in google.
What is Learning History?

When a registered user logs into the portal and looks for some information and reads it, the viewed link will be recorded by the system for future reference.

Accessible Content Based Collaborative Learning

Assume a new user completes his registration process and searches for a particular data. When the user gets his content ready for reading he will have an option to view top 5 or 10 users who have read the same content or updated the same content before, So the user can just click on the listed users and view their learning history.

If the user is new to the portal he/she will see the other user's learning history and get to know more about his learning path.

The algorithm at the backend should be intelligent enough to show a tree based structure{more similar to youtube} to the user, with which he/she will be traversing thru various contents ( Date Wise). In long run, it doesn't seem to be a feasible solution but we can truncate the tree structure to display the best linked contents and it is the responsibility of the algorithm.

Adaptive & Predictive Algorithm which is the backbone of the entire process, which will constantly watch each and every user and it will predict the user's learning scheme. This part can be discussed later.

If the user is interested in the learning path of other user, he/she can mark the user, so that there will be daily updates to the viewer on various link/content the marked user has traversed.
This will make the learning easy for new users and the purpose of learning will also be accomplished. If the user marks the best learner he/she will be accomplishing the term "learn how to learn".

This collaborative learning scheme of a user can be called as learning patterns. To measure the user and the content quality there are some simple inputs which can be derived from the portal itself, like user rating & content rating which is available in almost all the knowledge/learning portals.

The User will be allowed to mark more than one user and form a group. The Adaptive & Predictive Algorithm will try to combine all the user's contents/links to form a consolidated learning path. This part will let the user to have a defined path on learning and save more time and get improved expertise in a particular domain.
I believe this part will be the most critical when it comes to implementation.

Some More easy Learning...

The user can mark other users and get feeds to their systems which will make this more and more efficient.
The user will have graphical charts to see the learning curve of the other user.
The user can share his learning path to other users, this is more like del.icio.us.
The system itself can generate a best learning path and suggest the same to the users.

This forecast for e-Learning will drive good learning paths in a long run.

Users who read this link can give more ideas and I would like to improve this concept in a large scale. I have done some homework on this and I would like everybody to involve in this to make this as the future of learning and thus e-Learning.