Transformer-Based Multi-Hop Question Generation (Student Abstract)

Question generation is the parallel task of question answering, where given an input context and, optionally, answer, goal to generate a relevant and fluent natural language question. Although recent works on have experienced success by utilizing sequence-to-sequence models, there need for models handle increasingly complex contexts produce detailed questions. Multi-hop more challenging that aims questions connecting multiple facts from contexts. In this work, we apply transformer model multi-hop without any sentence-level supporting fact information. We utilize concepts proven effective in single-hop generation, including copy mechanism placeholder tokens. evaluate our model’s performance HotpotQA dataset using automated evaluation metrics, BLEU, ROUGE METEOR show improvement over previous work.