Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process

Guangming Huang, Yunfei Long, Cunjin Luo, Jiaxing Shen, Xia Sun

TL;DR

A Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA, and shows that PEI performs comparably to the state-of-the-art on HotpotQA.

Abstract

Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.

Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process

TL;DR

A Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA, and shows that PEI performs comparably to the state-of-the-art on HotpotQA.

Abstract

Pre-trained language models (PLMs) leverage chains-of-thought (CoT) to simulate human reasoning and inference processes, achieving proficient performance in multi-hop QA. However, a gap persists between PLMs' reasoning abilities and those of humans when tackling complex problems. Psychological studies suggest a vital connection between explicit information in passages and human prior knowledge during reading. Nevertheless, current research has given insufficient attention to linking input passages and PLMs' pre-training-based knowledge from the perspective of human cognition studies. In this study, we introduce a Prompting Explicit and Implicit knowledge (PEI) framework, which uses prompts to connect explicit and implicit knowledge, aligning with human reading process for multi-hop QA. We consider the input passages as explicit knowledge, employing them to elicit implicit knowledge through unified prompt reasoning. Furthermore, our model incorporates type-specific reasoning via prompts, a form of implicit knowledge. Experimental results show that PEI performs comparably to the state-of-the-art on HotpotQA. Ablation studies confirm the efficacy of our model in bridging and integrating explicit and implicit knowledge.
Paper Structure (19 sections, 3 equations, 3 figures, 5 tables)

This paper contains 19 sections, 3 equations, 3 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Chain-of-Thought Prompting
  4. Prompt-based Learning for Multi-hop QA
  5. Methodology
  6. Overview
  7. Type Prompter
  8. Knowledge Prompter
  9. Unified Prompter
  10. Experiments
  11. Dataset and Metrics
  12. Implementation Details
  13. Main Results and Analysis
  14. Evaluation of Robustness
  15. Ablation Studies
  16. ...and 4 more sections

Figures (3)

  • Figure 1: An example of the significance of implicit knowledge in reading comprehension.
  • Figure 2: The overview of our proposed PEI framework for multi-hop QA. The right green dashed block is the type prompter; the top blue dashed block refers to the knowledge prompter; and the bottom orange dashed block is the unified prompter.
  • Figure 3: The success rate (%) of five multi-hop QA models. $sub1$ denotes the first sub-question and $sub2$ is the second sub-question of corresponding question $q$. The results of DFGN, DecompRC, HGN and PCL are from deng2022prompt.