More Like this

1. Quantifying Causal Path-Specific Importance in Structural Causal Model

   https://doi.org/10.3390/computation11070133  

   Wang, Xiaoxiao ; Zhao, Minda ; Meng, Fanyu ; Liu, Xin ; Kong, Zhaodan ; Chen, Xin ( July 2023 , Computation)

   Path-specific effect analysis is a powerful tool in causal inference. This paper provides a definition of causal counterfactual path-specific importance score for the structural causal model (SCM). Different from existing path-specific effect definitions, which focus on the population level, the score defined in this paper can quantify the impact of a decision variable on an outcome variable along a specific pathway at the individual level. Moreover, the score has many desirable properties, including following the chain rule and being consistent. Finally, this paper presents an algorithm that can leverage these properties and find the k-most important paths with the highest importance scores in a causal graph effectively.

    

   more » « less
2. Commentary on “Causal Decision Making and Causal Effect Estimation Are Not the Same… and Why It Matters”

   https://doi.org/10.1287/ijds.2021.0010  

   McFowland, Edward ( January 2022 , INFORMS Journal on Data Science)
3. Text as Causal Mediators: Research Design for Causal Estimates of Differential Treatment of Social Groups via Language Aspects

   https://doi.org/10.18653/v1/2021.cinlp-1.2  

   Keith, Katherine ; Rice, Douglas ; O’Connor, Brendan ( January 2021 , First Workshop on Causal Inference & NLP (CI-NLP) at EMNLP 2021)
4. Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates

   https://doi.org/10.18653/v1/2020.acl-main.474  

   Keith, Katherine ; Jensen, David ; O’Connor, Brendan ( January 2020 , Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics)

   Many applications of computational social science aim to infer causal conclusions from non-experimental data. Such observational data often contains confounders, variables that influence both potential causes and potential effects. Unmeasured or latent confounders can bias causal estimates, and this has motivated interest in measuring potential confounders from observed text. For example, an individual’s entire history of social media posts or the content of a news article could provide a rich measurement of multiple confounders.Yet, methods and applications for this problem are scattered across different communities and evaluation practices are inconsistent.This review is the first to gather and categorize these examples and provide a guide to data-processing and evaluation decisions. Despite increased attention on adjusting for confounding using text, there are still many open problems, which we highlight in this paper. 

   more » « less
5. Modeling Document-level Causal Structures for Event Causal Relation Identification

   Gao, Lei ; Choubey, Prafulla Kumar ; Huang, Ruihong ( January 2019 , Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers))