Commonsense reasoning is important for many artificial intelligence (AI) tasks, from computer vision to natural language processing (NLP) and planning. It has also been a long-standing challenge in the field. In the last decade, there has been significant progress in AI owing to deep learning approaches which, in some cases, match or surpass human performance in tasks such as object recognition and machine translation. However, a lack of commonsense reasoning in deep learning models, which leads to brittle systems, is also acknowledged. In this presentation, I will discuss the recent efforts in commonsense knowledge representation and reasoning, with an emphasis on the study of the topic from an NLP perspective.
Related Materials: [Box-Folder]