GNET-QG: graph network for multi-hop question generation
No Thumbnail Available
Date
2024-08-28
Authors
Jamshidi, Samin
University of Lethbridge. Faculty of Arts and Science
Journal Title
Journal ISSN
Volume Title
Publisher
Lethbridge, Alta. : University of Lethbridge, Dept. of Computer Science
Abstract
Question generation involves crafting questions based on a given input context and accompanying answers. While recent advancements in sequence-to-sequence models have proven successful in generating natural language questions, there is an increasing need for models capable of handling more intricate contexts to produce detailed questions. Multi-hop question generation, which is more challenging, aims to establish connections between multiple facts from diverse input contexts to formulate questions. In our research, we study the utilization of a Graph Attention Network (GAT) and a BART model for multi-hop question generation, Our proposed model, is GNET-QG (Graph Network for Multi-Hop Question Generation). GNET-QG efficacy is assessed on the HotpotQA dataset using metrics such as METEOR, BLEU, and ROUGE, showcasing an enhancement over previous methodologies.
Description
Keywords
Question generation , Multi-hop question generation , Formulating questions , Natural language questions