ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning
This addresses the need for structured inferential knowledge in AI, enabling better commonsense reasoning for applications like natural language understanding, though it is incremental in building on existing resources.
The paper tackles the problem of commonsense reasoning by introducing ATOMIC, a dataset of 877k textual descriptions for if-then relations, and shows that neural models trained on it acquire commonsense capabilities, with multitask models achieving more accurate inference as measured by automatic and human evaluation.
We present ATOMIC, an atlas of everyday commonsense reasoning, organized through 877k textual descriptions of inferential knowledge. Compared to existing resources that center around taxonomic knowledge, ATOMIC focuses on inferential knowledge organized as typed if-then relations with variables (e.g., "if X pays Y a compliment, then Y will likely return the compliment"). We propose nine if-then relation types to distinguish causes vs. effects, agents vs. themes, voluntary vs. involuntary events, and actions vs. mental states. By generatively training on the rich inferential knowledge described in ATOMIC, we show that neural models can acquire simple commonsense capabilities and reason about previously unseen events. Experimental results demonstrate that multitask models that incorporate the hierarchical structure of if-then relation types lead to more accurate inference compared to models trained in isolation, as measured by both automatic and human evaluation.