ROCK: Causal Inference Principles for Reasoning about Commonsense Causality

TL;DR

ROCK introduces a causal inference framework for commonsense reasoning that leverages temporal signals and propensity scores to better identify plausible causes and effects in natural language, addressing limitations of current deep learning approaches.

Contribution

It is the first to adapt the potential-outcomes causal framework to commonsense reasoning, using temporal signals and propensity scores for better confounder balancing.

Findings

ROCK demonstrates strong zero-shot CCR capabilities.

The framework effectively balances confounding effects.

It outperforms existing methods in plausibility assessments.

Abstract

Commonsense causality reasoning (CCR) aims at identifying plausible causes and effects in natural language descriptions that are deemed reasonable by an average person. Although being of great academic and practical interest, this problem is still shadowed by the lack of a well-posed theoretical framework; existing work usually relies on deep language models wholeheartedly, and is potentially susceptible to confounding co-occurrences. Motivated by classical causal principles, we articulate the central question of CCR and draw parallels between human subjects in observational studies and natural languages to adopt CCR to the potential-outcomes framework, which is the first such attempt for commonsense tasks. We propose a novel framework, ROCK, to Reason O(A)bout Commonsense K(C)ausality, which utilizes temporal signals as incidental supervision, and balances confounding effects using temporal propensities that are analogous to propensity scores. The ROCK implementation is modular and zero-shot, and demonstrates good CCR capabilities.