Supervised data mining is when you have a predefined target variable you want to predict or explain. For example, let's say you have a set of clients and you want to predict the likelihood of them cancelling their current contract.
That's a typical supervised data mining problem.
The thing with supervised data mining is that you need quality data on the target variable to train your model. In our example, you need a history of clients who cancelled and clients who did not cancel their contracts to "feed" the model and train it.