I create an application with some social interaction between users. Users will have ratings that determines your recommendations - users are recommended to users with the same rating.Total rating of user is the sum of the partial ratings multiplied by their coefficients of significance.

$totalRating = \sum_{i=0} partialRating_i*coefficient_i$

But only the total rating is used in the final for recommendations.

Each partial rating has a coefficient of significance and a decreasing time interval.

Examples of partial ratings:

  • online status. When the user logs into the application. this rating rise to 100, after 5 hours(decreasing time interval) of inactivity it decreases to 75, after 12 hours of inactivity - 50, etc.
  • boost for beginners. When the user creates his profile, this rating is equals to 100 and decreases every 2 days by 25 points, until it drops to zero.
  • activity in application. The more a user interacts with other users, the greater his rating. As with other ratings, it drops to zero after a certain period of inactivity.

So, the main headache - how to calculate the most suitable coefficients of significance and decreasing time intervals?

I understand that these parameters will be selected rather based on my preferences. But is there any method that can somehow help choose these parameters?


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.