Dense-ATOMIC: Construction of Densely-connected and Multi-hop Commonsense Knowledge Graph upon ATOMIC

10/14/2022
by   Xiangqing Shen, et al.
0

ATOMIC is a large-scale commonsense knowledge graph (CSKG) containing everyday if-then knowledge triplets, i.e., head event, relation, tail event. The one-hop annotation manner made ATOMIC a set of independent bipartite graphs, which ignored the numerous missing links between events in different bipartite graphs and consequently caused shortcomings in knowledge coverage and multi-hop reasoning. To address these issues, we propose a CSKG completion approach by training a relation prediction model based on a set of existing triplets, and infer the missing links on ATOMIC. On this basis, we construct Dense-ATOMIC, a densely-connected and multi-hop commonsense knowledge graph. The experimental results on an annotated dense subgraph demonstrate the effectiveness of our CSKG completion approach upon ATOMIC. The evaluation on a downstream commonsense reasoning task also proves the advantage of Dense-ATOMIC against conventional ATOMIC.

READ FULL TEXT
research
09/24/2020

Language Generation with Multi-Hop Reasoning on Commonsense Knowledge Graph

Despite the success of generative pre-trained language models on a serie...
research
03/21/2022

Relevant CommonSense Subgraphs for "What if..." Procedural Reasoning

We study the challenge of learning causal reasoning over procedural text...
research
08/13/2020

Commonsense Knowledge Graph Reasoning by Selection or Generation? Why?

Commonsense knowledge graph reasoning(CKGR) is the task of predicting a ...
research
10/31/2018

ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning

We present ATOMIC, an atlas of everyday commonsense reasoning, organized...
research
10/28/2021

SMORE: Knowledge Graph Completion and Multi-hop Reasoning in Massive Knowledge Graphs

Knowledge graphs (KGs) capture knowledge in the form of head–relation–ta...
research
04/20/2021

Identify, Align, and Integrate: Matching Knowledge Graphs to Commonsense Reasoning Tasks

Integrating external knowledge into commonsense reasoning tasks has show...
research
08/29/2019

NarrativeTime: Dense High-Speed Temporal Annotation on a Timeline

We present NarrativeTime, a new timeline-based annotation scheme for tem...

Please sign up or login with your details

Forgot password? Click here to reset