Commonsense Knowledge
Commonsense knowledge research aims to equip artificial intelligence systems with the everyday understanding that humans possess implicitly. Current research focuses on integrating commonsense into various AI tasks, such as robotics, dialogue systems, and image understanding, often leveraging large language models, graph convolutional networks, and knowledge graphs to represent and reason with this knowledge. This work is crucial for improving AI's reliability, explainability, and ability to interact naturally with humans, ultimately leading to more robust and user-friendly AI applications across numerous domains. The development of new benchmarks and datasets is also a significant focus, enabling more rigorous evaluation and comparison of different approaches.
Papers
PseudoReasoner: Leveraging Pseudo Labels for Commonsense Knowledge Base Population
Tianqing Fang, Quyet V. Do, Hongming Zhang, Yangqiu Song, Ginny Y. Wong, Simon See
Extracting Cultural Commonsense Knowledge at Scale
Tuan-Phong Nguyen, Simon Razniewski, Aparna Varde, Gerhard Weikum
MICO: A Multi-alternative Contrastive Learning Framework for Commonsense Knowledge Representation
Ying Su, Zihao Wang, Tianqing Fang, Hongming Zhang, Yangqiu Song, Tong Zhang
Probing Commonsense Knowledge in Pre-trained Language Models with Sense-level Precision and Expanded Vocabulary
Daniel Loureiro, Alípio Mário Jorge
CIKQA: Learning Commonsense Inference with a Unified Knowledge-in-the-loop QA Paradigm
Hongming Zhang, Yintong Huo, Yanai Elazar, Yangqiu Song, Yoav Goldberg, Dan Roth