Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Survey on Neural Question Generation: Methods, Applications, and Prospects

Shasha Guo, Lizi Liao, Cuiping Li, Tat-Seng Chua

TL;DR

The surveyed work articulates a comprehensive landscape of Neural Question Generation (NQG) across structured, unstructured, and hybrid data modalities, detailing problem formulation, datasets, metrics, and applications. It surveys three architecture families for structured QG (Traditional Seq2Seq, Graph2Seq, and Pre-trained Seq2Seq), and extends to unstructured QG (text and visual) with corresponding modeling paradigms, including graph-based and PLM/LLM-driven approaches. The paper also discusses hybrid paradigms that fuse multiple modalities, and points to emergent trends such as in-context learning and prompt-based KBQG with LLMs. Collectively, the work provides a foundation for future research in proactive, multi-modal, and controllable QG, while calling for improved evaluation metrics and cross-domain benchmarks to better capture question quality and usefulness.

Abstract

In this survey, we present a detailed examination of the advancements in Neural Question Generation (NQG), a field leveraging neural network techniques to generate relevant questions from diverse inputs like knowledge bases, texts, and images. The survey begins with an overview of NQG's background, encompassing the task's problem formulation, prevalent benchmark datasets, established evaluation metrics, and notable applications. It then methodically classifies NQG approaches into three predominant categories: structured NQG, which utilizes organized data sources, unstructured NQG, focusing on more loosely structured inputs like texts or visual content, and hybrid NQG, drawing on diverse input modalities. This classification is followed by an in-depth analysis of the distinct neural network models tailored for each category, discussing their inherent strengths and potential limitations. The survey culminates with a forward-looking perspective on the trajectory of NQG, identifying emergent research trends and prospective developmental paths. Accompanying this survey is a curated collection of related research papers, datasets and codes, systematically organized on Github, providing an extensive reference for those delving into NQG.

A Survey on Neural Question Generation: Methods, Applications, and Prospects

TL;DR

The surveyed work articulates a comprehensive landscape of Neural Question Generation (NQG) across structured, unstructured, and hybrid data modalities, detailing problem formulation, datasets, metrics, and applications. It surveys three architecture families for structured QG (Traditional Seq2Seq, Graph2Seq, and Pre-trained Seq2Seq), and extends to unstructured QG (text and visual) with corresponding modeling paradigms, including graph-based and PLM/LLM-driven approaches. The paper also discusses hybrid paradigms that fuse multiple modalities, and points to emergent trends such as in-context learning and prompt-based KBQG with LLMs. Collectively, the work provides a foundation for future research in proactive, multi-modal, and controllable QG, while calling for improved evaluation metrics and cross-domain benchmarks to better capture question quality and usefulness.

Abstract

In this survey, we present a detailed examination of the advancements in Neural Question Generation (NQG), a field leveraging neural network techniques to generate relevant questions from diverse inputs like knowledge bases, texts, and images. The survey begins with an overview of NQG's background, encompassing the task's problem formulation, prevalent benchmark datasets, established evaluation metrics, and notable applications. It then methodically classifies NQG approaches into three predominant categories: structured NQG, which utilizes organized data sources, unstructured NQG, focusing on more loosely structured inputs like texts or visual content, and hybrid NQG, drawing on diverse input modalities. This classification is followed by an in-depth analysis of the distinct neural network models tailored for each category, discussing their inherent strengths and potential limitations. The survey culminates with a forward-looking perspective on the trajectory of NQG, identifying emergent research trends and prospective developmental paths. Accompanying this survey is a curated collection of related research papers, datasets and codes, systematically organized on Github, providing an extensive reference for those delving into NQG.
Paper Structure (18 sections, 1 equation, 1 figure, 1 table)

This paper contains 18 sections, 1 equation, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Background
  3. Problem Formulation
  4. Popular Datasets
  5. Evaluation
  6. Applications
  7. Structured Neural Question Generation
  8. Traditional Seq2Seq Models
  9. Graph2Seq Models
  10. Pre-trained Seq2Seq Models
  11. Unstructured Neural Question Generation
  12. TQG
  13. Traditional Seq2Seq Models
  14. Graph-based Models
  15. Pre-trained Seq2Seq Models
  16. ...and 3 more sections

Figures (1)

  • Figure 1: The taxonomy of NQG. We classify NQG into three types based on input modalities: Structured NQG, which deals with structured data; Unstructured NQG, which handles unstructured data; and Hybrid NQG, which integrates both structured and unstructured data.