What can we infer about people by looking at their social network profiles? 
Or by seeing their responses in a test when we don't even know the right answers?

We'll see how machine learning techniques allow us to mine social network data to gain deep knowledge about people, including their demographics, personality and emotions. 
Similarly, we'll see how a probabilistic graphical model can allow a teacher to determine students' level of intelligence just by seeing how they answer a test for which the right answers are not known.