1
$\begingroup$

Given the (cosine) similarity score of top 100 neighbors of every item, how do I predict ratings for unrated items? Please explain in simple terms.

Item
1 260 0.577305 780 0.5655413 1210 0.5529503 3114 0.5425038 1270 .....

2 367 0.5202925 364 0.5093084 500 0.5082204 586 0.4978301 480 ......

. . .

$\endgroup$
1
$\begingroup$

You should at first calculate the similarity between co-rated items (items which both active user and another user rated them). then you can predict the rate of active user to the unrated item. one of the methods for this work is Slope One

$\endgroup$

Your Answer

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

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