0
$\begingroup$

Let's say a User was Interested in football, so we were recommending him posts about football after few days or weeks he starts watching Baseball.

How'd we go about building a recommender system that also takes recency into consideration Or Let's say that it knows when a user was interested in what

$\endgroup$
1
  • 2
    $\begingroup$ Just let the age of the item be one of the features then. Note that this is different from the contextual case where the actual time of the day is a consideration. Make sure you are not confusing them. Welcome to the site! $\endgroup$
    – Emre
    Nov 26, 2017 at 22:48

1 Answer 1

1
$\begingroup$

Use a diminishing sample weight where the newest sample = 1 and older samples get lower weights. This way if a user changes his interests over time, your recommendations will change accordingly.

$\endgroup$
1
  • $\begingroup$ alright man would try to do that but I'm keeping the question open for suggestion if this worked I'll make sure to close it $\endgroup$ Nov 27, 2017 at 13:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.