Huang, Guangming and Long, Yunfei and Luo, Cunjin and Shen, Jiaxing and Sun, Xia (2024) Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process. In: The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, 2024-05-20 - 2024-05-25, Torino, Italy.
Huang, Guangming and Long, Yunfei and Luo, Cunjin and Shen, Jiaxing and Sun, Xia (2024) Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process. In: The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, 2024-05-20 - 2024-05-25, Torino, Italy.
Huang, Guangming and Long, Yunfei and Luo, Cunjin and Shen, Jiaxing and Sun, Xia (2024) Prompting Explicit and Implicit Knowledge for Multi-hop Question Answering Based on Human Reading Process. In: The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, 2024-05-20 - 2024-05-25, Torino, Italy.
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 input passages and human prior knowledge during reading. Nevertheless, current research has given less attention to linking input passages and PLMs’ pre-training-based knowledge derived from human reading processes. 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, using it to elicit implicit knowledge through unified prompt reasoning. Additionally, 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 effectiveness of our model in bridging and incorporating explicit and implicit knowledge.
Item Type: | Conference or Workshop Item (Paper) |
---|---|
Uncontrolled Keywords: | Implicit Knowledge; Multi-hop QA; Prompt |
Divisions: | Faculty of Science and Health Faculty of Science and Health > Computer Science and Electronic Engineering, School of |
SWORD Depositor: | Unnamed user with email elements@essex.ac.uk |
Depositing User: | Unnamed user with email elements@essex.ac.uk |
Date Deposited: | 02 Aug 2024 10:49 |
Last Modified: | 02 Aug 2024 10:50 |
URI: | http://repository.essex.ac.uk/id/eprint/37847 |
Available files
Filename: 2024.lrec-main.1154.pdf
Licence: Creative Commons: Attribution-Noncommercial 4.0