Scalable Multi-Hop Relational Reasoning for Knowledge-Aware Question Answering

By Yanlin Feng, Xinyue Chen, Bill Yuchen Lin, Peifeng Wang, Jun Yan, Xiang Ren. 2020

288 citations recorded by the source corpus.

Existing work on augmenting question answering (QA) models with external knowledge (e.g., knowledge graphs) either struggle to model multi-hop relations efficiently, or lack transparency into the model's prediction rationale. In this paper, we propose a novel knowledge-aware approach that equips pre-trained language models (PTLMs) with a multi-hop relational reasoning module, named multi-hop graph relation network (MHGRN). It performs multi-hop, multi-relational reasoning over subgraphs extracted from external knowledge graphs. The proposed reasoning module unifies path-based reasoning methods and graph neural networks to achieve better interpretability and scalability. We also empirically show its effectiveness and scalability on CommonsenseQA and OpenbookQA datasets, and interpret its behaviors with case studies.

Subfields: Artificial Intelligence, Question Answering & Reading Comprehension, Knowledge Graphs & Representation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Question Answering & Reading Comprehension

In Knowledge Graphs & Representation

Citations in the corpus

Enhance Multimodal Transformer With External Label And In-Domain Pretrain: Hateful Meme Challenge Winning Solution (2020, background): “Moreover, the attention GNN used in MHGRN makes it easier to scale the model to a larger graph, which may be a critical part of the problem when dealing with the subtle relationship between key entities inside a meme.”
Transform-Retrieve-Generate: Natural Language-Centric Outside-Knowledge Visual Question Answering (2022, background): “There are mainly two streams of approaches, namely knowledge graph-based question answering [10, 32, 38, 52, 59, 65] and knowledge retrieval-based question answering [5, 7, 19, 29, 44, 45, 60, 62].”
Making Language Models Better Reasoners with Step-Aware Verifier (2022, background): “Much research in the literature enhances the reasoning capabilities of machine learning systems by exploiting symbolic systems, including knowledge graphs (Mihaylov and Frank, 2018; Bauer et al., 2018; Kundu et al., 2019; Wang et al., 2019; Lin et al., 2019; Ding et al., 2019; Feng et al., 2020; Wan”
Knowledge-Augmented Methods for Natural Language Processing (2023, methodology): “For themultiplechoice QA setting, we introduce a set of knowledge-augmented neuro-symbolic methods: KagNet [23], HyKAS [30], MHGRN [8], and HybridGN [43].”
GreaseLM: Graph REASoning Enhanced Language Models for Question Answering (2022, methodology): “Others reverse the ﬂow of information and use a representation of the text ( e.g. , ﬁnal layer of LM) to provide an augmentation to a graph reasoning model over an extracted KG for the example ( e.g. , MHGRN, Feng et al., 2020; Lv et al., 2020).”
GraDA: Graph Generative Data Augmentation for Commonsense Reasoning (2022, background): “Bauer et al. (2018); Lin et al. (2019); De Cao et al. (2019); Feng et al. (2020) and Lv et al. (2020) extract relevant multi-hop relational commonsense from knowledge graphs and show significant improvements over models that operate solely on text.”
Large Language Models as Analogical Reasoners (2023, background): “…training (Wu et al., 2021; Yasunaga et al., 2022b; Lightman et al., 2023; Moor et al., 2023), retrieving structured knowledge (Lin et al., 2019; Feng et al., 2020; Zhang et al., 2022a; Yasunaga et al., 2021; 2022a; Xie et al., 2022), and incorporating external modules for reasoning such as…”
Differentiable Open-Ended Commonsense Reasoning (2020, methodology): “For example, KagNet (Lin et al., 2019) and MHGRN (Feng et al., 2020) use an external commonsense knowledge graph to retrieve the connections between the question and each given candidate, and then use them as structural priors to individually score each choice.”
Graph Retrieval-Augmented Generation: A Survey (2024, unknown): “OpenCSR [45] divides the retrieval process into two stages.”
Deep Bidirectional Language-Knowledge Graph Pretraining (2022, unknown): “Here we experimented with an alternative pretraining method that drops the graph structure: we convert triplets in the local KG into sentences using a template (Feng et al. 2020), append them to the main text input, and perform vanilla MLM pre-training.”
On the Advance of Making Language Models Better Reasoners (2022, background): “…systems, including knowledge graphs (Mihaylov and Frank, 2018; Bauer et al., 2018; Kundu et al., 2019; Wang et al., 2019; Lin et al., 2019; Ding et al., 2019; Feng et al., 2020; Wang et al., 2022b), or question taxonomies (Dua et al., 2019; Andor et al., 2019; Hu et al., 2019b; Wang et al., 2022a).”
Retrieval Augmented Visual Question Answering with Outside Knowledge (2022, background): “, 2020), knowledge-graphs (KGs) (Lin et al., 2019; Feng et al., 2020; Lv et al., 2020; Saffari et al., 2021) or document collections (Chen et al.”
Think about it! Improving defeasible reasoning by first modeling the question scenario. (2021, methodology): “Encoding Graph Representations Several existing methods use graphs as an additional input for commonsense reasoning (Sun et al., 2018; Lin et al., 2019; Lv et al., 2020; Feng et al., 2020; Bosselut et al., 2021; Ma et al., 2021; Kapani-pathi et al., 2020).”
QA-GNN: Reasoning with Language Models and Knowledge Graphs for Question Answering (2021, methodology): “We first encode the QA context using an LM, and retrieve a KG subgraph following prior works (Feng et al., 2020).”
Unifying Large Language Models and Knowledge Graphs: A Roadmap (2023, methodology): “Knowledge graph question answering (KGQA) aims to find answers to natural language questions based on the structured facts stored in knowledge graphs [233], [234].”
Connecting the Dots: A Knowledgeable Path Generator for Commonsense Question Answering (2020, result): “Parts of the results for baselines are reported from our another work (Feng et al., 2020).”
Exploiting Reasoning Chains for Multi-hop Science Question Answering (2021, methodology): “Comparison Methods: For fair comparison, we compare CGR with recently published methods that use similar power of pretrained models, including five textual knowledge based methods: AristoRoBERTa (AllenAI, 2019), KF-SIR (Banerjee and Baral, 2020), FreeLB (Zhu et al., 2020), DPR (Karpukhin et al., 202”