Entropy
Overview
Entropy is a measure of uncertainty, or the amount of information to reduce the uncertainty. The formal definition is as follows.
Given a discrete random variable X, with possible outcomes x_{1},...,x_{n}, which occur with probability \mathrm {P} (x_{1}),...,\mathrm {P} (x_{n}), the entropy of X is formally defined as: $$ \mathrm {H} (X)=-\sum_{i=1}^{n}{\mathrm {P} (x_i)\log \mathrm {P} (x_i)} $$
You can think entropy as something related to the encoding of all the outcomes. We use 0 and 1 to encode them. The minimum average length of the all the encodings is the entropy.
The more outcomes a random variable has, the larger the entropy is. The more unbalance the probability distribution is, the smaller the entropy is.
The Logarithm
You can think the -\log \mathrm {P} (x_i) as a measurement for the encoding length of that specific result X=x_i. For example, if every outcome has a same probability of \frac{1}{8}, the encoding length with binary bits should be -\log_2\frac{1}{8} = 3.
An Expectation Perspective
Entropy is an expectation of the encoding length of different outcomes.