ROCK: Causal Inference Principles for Reasoning about Commonsense Causality

Jiayao Zhang, Hongming Zhang, Weijie Su, Dan Roth
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:26750-26771, 2022.

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.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-zhang22am, title = {{ROCK}: Causal Inference Principles for Reasoning about Commonsense Causality}, author = {Zhang, Jiayao and Zhang, Hongming and Su, Weijie and Roth, Dan}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {26750--26771}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/zhang22am/zhang22am.pdf}, url = {https://proceedings.mlr.press/v162/zhang22am.html}, 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.} }
Endnote
%0 Conference Paper %T ROCK: Causal Inference Principles for Reasoning about Commonsense Causality %A Jiayao Zhang %A Hongming Zhang %A Weijie Su %A Dan Roth %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-zhang22am %I PMLR %P 26750--26771 %U https://proceedings.mlr.press/v162/zhang22am.html %V 162 %X 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.
APA
Zhang, J., Zhang, H., Su, W. & Roth, D.. (2022). ROCK: Causal Inference Principles for Reasoning about Commonsense Causality. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:26750-26771 Available from https://proceedings.mlr.press/v162/zhang22am.html.

Related Material