Model based approach prediction: how?

[alucard001]

Hi, I am currently reading this book and when I am in Chapter 6 in your code, I have a question on the model based approach:

Now when we have evaluated knn to have RMSE of 0.97, now what?

What I mean is: now I love to use knn, I want to do some prediction, how?

How to do it in terms of item-based?
How to do it in terms of user-based?

Can you please give us some example of how it is actually used?

Actually, I read quite a lot of books on machine learning, and they ALL stop at the evaluation part. I don't know why can we just move 1 step further to illustrate how to do the prediction, as it is equally important to evaluation, but most books are missing this (how to do prediction) part.

Thank you.