It is called Adaptive Boosting due to the fact that the weights are re-assigned to each instance, with higher weights being assigned to instances that are not correctly classified - therefore it ‘adapts’.