chunked/nltk_chunking/tasks_document_question_answering/chunk_1.txt

In other words, the problem is treated as extractive question answering: given the context, extract which piece
of information answers the question.