Complex Reasoning over Logical Queries on Commonsense Knowledge Graphs
This addresses data scarcity for training models on complex event interactions in commonsense reasoning, though it is incremental as it builds on existing knowledge graphs and verbalization methods.
The paper tackles the challenge of event commonsense reasoning by creating COM2, a dataset of multi-hop logical queries from a commonsense knowledge graph, which improves language models' complex reasoning ability, leading to significant gains in zero-shot performance for question answering and generative tasks.
Event commonsense reasoning requires the ability to reason about the relationship between events, as well as infer implicit context underlying that relationship. However, data scarcity makes it challenging for language models to learn to generate commonsense inferences for contexts and questions involving interactions between complex events. To address this demand, we present COM2 (COMplex COMmonsense), a new dataset created by sampling multi-hop logical queries (e.g., the joint effect or cause of both event A and B, or the effect of the effect of event C) from an existing commonsense knowledge graph (CSKG), and verbalizing them using handcrafted rules and large language models into multiple-choice and text generation questions. Our experiments show that language models trained on COM2 exhibit significant improvements in complex reasoning ability, resulting in enhanced zero-shot performance in both in-domain and out-of-domain tasks for question answering and generative commonsense reasoning, without expensive human annotations. Code and data are available at https://github.com/tqfang/complex-commonsense-reasoning.