Thursday, August 4, 2011

Recommendation Systems

The Recommendation Systems (RS) has been termed as the next step
in the evolution of Artificial Intelligence + search systems on the web ,
together being termed as intelligent search systems
that enable a user to search a query on a website conveniently and flexibly ,
without thinking too hard.

Sometimes RS being the "best friend" advisor in case of high indecisiveness .
( P.S. sure to take it when you go shopping next time ? )



It also evolves the "feel" organ of computer ,
  • dynamically i.e. via real time algorithms OR
  • via database support systems that work on analyzing pre-stored information regarding the user's search patterns and preferences.

Based on social networking and surfing concepts of
  • rating an item or user OR
  • following a "friend" or member OR
  • recommendations based on the user's previous search and selections OR
  • choices based on interests OR
  • popular general vote OR
  • comments/feedback of friends/other users


Also includes the "convolution" of different features
combined in a complimentary way such as
to develop a feasible user search algorithm depending upon the requirements .


Classified by the types :
  • search based ,
  • cateogary based ,
  • info filtering ,
  • classifiers ,
  • clusters ,
  • collaborative filtering and many others ,

being used extensively on the World Wide Web .

More to follow up on RS soon ... happy reading!

No comments:

Post a Comment