Generating Questions for Reading Comprehension using Coherence Relations

摘要

We propose a technique for generating complex reading comprehension questions from a discourse that are more useful than factual ones derived from assertions. Our system produces a set of general-level questions using coherence relations. These evaluate comprehension abilities like comprehensive analysis of the text and its structure, correct identification of the author's intent, thorough evaluation of stated arguments; and deduction of the high-level semantic relations that hold between text spans. Experiments performed on the RST-DT corpus allow us to conclude that our system possesses a strong aptitude for generating intricate questions. These questions are capable of effectively assessing student interpretation of text.