Classification is a typical task of supervised machine learning. Its aim is to predict the class of certain categories. In order to do so, an algorithm is trained that is able to assign label to any type of data (e.g. text, image). When do you face an issue of classification? When you have pre-determined categories which you want to put certain items into. E.g. you intend to assign incoming emails to the competent team that has the responsibility to deal with them. It is needed to bear in mind that you do not necessarily have to know the features of the categories, since our smart automatic classifier is able to find them out. E.g. regarding performance evaluation, knowing the scale of evaluation (e.g. from 1 to 5) is satisfactory as the distinguishing features are invented by the classifier.
For more about the issue at hand, read our project on classification.