COMET: Commonsense Transformers for Automatic Knowledge Graph Construction

We present the first comprehensive study on automatic knowledge base construction for two prevalent commonsense knowledge graphs: ATOMIC Contrary to many conventional KBs that store knowledge with canonical templates, commonsense KBs only store loosely structured open-text descriptions of knowledge. We posit that an important step toward automatic commonsense completion is the development of generative models of commonsense knowledge, and propose COMmonsEnse Transformers (COMET ) that learn to generate rich and diverse commonsense descriptions in natural language. Despite the challenges of commonsense modeling, our investigation reveals promising results when implicit knowledge from deep pre-trained language…

• NS
  National Science Foundation

  Awards: 1256082, DGE-1256082, 1524371, 1714566, 1525251
• DA
  Defense Advanced Research Projects Agency
• S
  Samsung
• AR
  Army Research Office

  Awards: W911NF-15-1-, W911NF