Last week I had the opportunity to provide a talk on the challenges on textual data stream classification at the Université de Rouen, in France. This talk discussed the types of concept drift that may coexist in textual data streams and also the techniques we are developing to address them.