In this issue: we discuss the concept of question-answering models; we overview a paper in which the Google Research team presents a new dataset for training and evaluating question-answering systems; we evaluate DeepPavlov, an open-source framework for advanced NLU methods, including question-answering