In this framework the problem is selected in a manner that would maximize student's knowledge, the information that she/he gets.This is formulated as: $s^{*} = argmax [p_sn_s^+ + (1-p_s)n_s^-]$

Here $p_s$ is the probability that the user will solve the problem s, if $n_s^+$ is the amount of problems that the user can solve correctly and $n_s^-$ how many problems user cannot solve.

The $p_s$ is estimated simply by: $p_s = \frac {N^+} {N^+ + N^-}$