Big data is a challenge because it can lead to information overload. Data was gathered and looked at, but then no action was taken. Most of our users want the data analyzed and then rolled up and packaged for them. Challenge is how to make it simple enough to be consumed easily while not losing the value.
Bias can lead to data misanalyzed. Seeing so many data points that it is easy to let your own world view affect your perspective. It may lead to seeing trends that don’t exist.
Big data projects should be seen as agile as possible. Define a very modest objective and strive to attain it first. It should deliver tangible results and then you can build on it in a positive way. It needs to be actionable so that it provides value.
Wow, this is the first session that I am finding is fairly dry and boring. So far I’ve enjoyed every session I have attended. But this one the speakers are very dry and dull. One of the panelists literally almost fell asleep in front of everyone. Now that is funny!
Data driven education is an early area of research. It can lead to personalized education, and how someone learns. The trick is to find how a student learns best in a quick automated way. That is ideally the future many years from now. This is going to take some time,
Dan at a Conference 
Same issue in Security. retention of data is also becoming an issue.