What is the model parameters in Naive Bayes?

Question

I just lead the Naive Bayes learning, the form is

$$ P(y, x_1, \dotsc, x_n) = p(y) \prod_{i=1}^n p(x_i \mid y). $$

In this lecture, it says

Each factor $ p(x_i \mid y) $ can be completely described by a small number of parameters (4 parameters with 2 degrees of freedom to be exact). The entire distribution is parametrized by O(n) parameters, which we can tractably estimate from data and make predictions.

My question is

1. What is the model parameters for Bayes learning.
2. How to understand: $ p(x_i \mid y) $ can be completely described by a small number of parameters (4 parameters with 2 degrees of freedom to be exact)?

Thanks.