Question Generation Using Sequence-to-Sequence Model with Semantic Role Labels

Automatic generation of questions from text has gained increasing attention due to its useful applications. We propose a novel question generation method that combines the benefits of rule-based and neural sequence-to-sequence (Seq2Seq) models. The proposed method can automatically generate multiple questions from an input sentence covering different views of the sentence as in rule-based methods, while more complicated "rules" can be learned via the Seq2Seq model. The method utilizes semantic role labeling (SRL) used in rule-based methods to convert training examples into their semantic representations, and then trains a sequence-to-sequence model over the semantic representations. Our extensive experiments on three real-world data sets show that the proposed method significantly improves the state-of-the-art neural question generation approaches in terms of both automatic and human evaluation measures. Moreover, we extend our proposed approach to a paragraph-level SRL-based method and evaluate it on two data sets. Through both automatic and human evaluations, we show that our proposed framework remarkably improves its Seq2Seq counterparts.