Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond

Feder, Amir; Keith, Katherine A.; Manzoor, Emaad; Pryzant, Reid; Sridhar, Dhanya; Wood-Doughty, Zach; Eisenstein, Jacob; Grimmer, Justin; Reichart, Roi; Roberts, Margaret E.; Stewart, Brandon; Veitch, Victor; Yang, Diyi

doi:10.48550/arxiv.2109.00725

Cited by 31 publications

(35 citation statements)

References 80 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our work in no way seeks to diminish the active efforts of folks and social movements that continue, today and that have for decades, striven to advocate for the rights of, change the biased social perceptions towards, and champion the values of equality of traditionally marginalized populations and their lived experiences. One possibility for moving closer to truly causal studies is through direct participatory work or using causal inference techniques (Feder et al, 2021) to examine how attitudes influence the word selection or how reading particular uses influence the person's attitude or interpretation of the passage.…”

Section: Discussionmentioning

confidence: 99%

Using Sociolinguistic Variables to Reveal Changing Attitudes Towards Sexuality and Gender

CH-Wang¹,

Jurgens²

2021

Preprint

View full text Add to dashboard Cite

Individuals signal aspects of their identity and beliefs through linguistic choices. Studying these choices in aggregate allows us to examine large-scale attitude shifts within a population. Here, we develop computational methods to study word choice within a sociolinguistic lexical variable-alternate words used to express the same concept-in order to test for change in the United States towards sexuality and gender. We examine two variables: i) referents to significant others, such as the word "partner" and ii) referents to an indefinite person, both of which could optionally be marked with gender. The linguistic choices in each variable allow us to study increased rates of acceptances of gay marriage and gender equality, respectively. In longitudinal analyses across Twitter and Reddit over 87M messages, we demonstrate that attitudes are changing but that these changes are driven by specific demographics within the United States. Further, in a quasi-causal analysis, we show that passages of Marriage Equality Acts in different states are drivers of linguistic change.

show abstract

Section: Discussionmentioning

confidence: 99%

Using Sociolinguistic Variables to Reveal Changing Attitudes Towards Sexuality and Gender

CH-Wang¹,

Jurgens²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, we note that the term "sensitivity" is also used with other different meanings in the literature. Sensitivity test (Feder et al, 2021), e.g. Counterfactually Augmented Data (CAD) (Kaushik et al, 2019), evaluates the extent that models use spurious features to make predictions by injecting minimally label-flipping perturbations on target features.…”

Section: Related Workmentioning

confidence: 99%

“…Despite the success of deep neural models on many Natural Language Processing (NLP) tasks (Liu et al, 2016;Devlin et al, 2019;Liu et al, 2019b), recent work has discovered that these models rely excessively on spurious features, making the right predictions for the wrong reasons (Gururangan et al, 2018;McCoy et al, 2019;Wang & Culotta, 2020). Neural NLP models learn correlations but not causation from training data (Feder et al, 2021) and thus they are easily fooled by spurious correlation: prediction rules that work for the majority examples but do not hold in general (Tu et al, 2020). For example, BERT (Devlin et al, 2019) only achieves an accuracy less than 10% on a challenge test set HANS (McCoy et al, 2019) for MNLI (Williams et al, 2018) where spurious correlation disappears.…”

Section: Introductionmentioning

confidence: 99%

Interpreting the Robustness of Neural NLP Models to Textual Perturbations

Zhang¹,

Pan²,

Tan³

et al. 2021

Preprint

View full text Add to dashboard Cite

Recent work finds modern natural language processing (NLP) models relying on spurious features for prediction. Mitigating such effects is thus important. Despite this need, there is no quantitative measure to evaluate or compare the effects of different forms of spurious features in NLP. We address this gap in the literature by quantifying model sensitivity to spurious features with a causal estimand, dubbed CENT, which draws on the concept of average treatment effect from the causality literature. By conducting simulations with four prominent NLP models -TextRNN, BERT, RoBERTa and XLNet -we rank the models against their sensitivity to artificial injections of eight spurious features. We further hypothesize and validate that models that are more sensitive to a spurious feature will be less robust against perturbations with this feature during inference. Conversely, data augmentation with this feature improves robustness to similar perturbations. We find statistically significant inverse correlations between sensitivity and robustness, providing empirical support for our hypothesis.

show abstract

“…The outcome, however, may be confounded by legal representation, which exists only for a subset of cases and may matter substantially in whether an appeal is taken and succeeds. Instead of inferring spurious correlations, it will be important to identify deconfounded causal lexicons [46,23] in the SDM setting. FORMULATION.…”

Section: Causal Inferencementioning

confidence: 99%

Beyond Ads: Sequential Decision-Making Algorithms in Law and Public Policy

Henderson¹,

Chugg²,

Anderson³

et al. 2021

Preprint

View full text Add to dashboard Cite

We explore the promises and challenges of employing sequential decision-making algorithms -such as bandits, reinforcement learning, and active learning -in the public sector. While such algorithms have been heavily studied in settings that are suitable for the private sector (e.g., online advertising), the public sector could greatly benefit from these approaches, but poses unique methodological challenges for machine learning. We highlight several applications of sequential decision-making algorithms in regulation and governance, and discuss areas for further research which would enable them to be more widely applicable, fair, and effective. In particular, ensuring that these systems learn rational, causal decisionmaking policies can be difficult and requires great care. We also note the potential risks of such deployments and urge caution when conducting work in this area. We hope our work inspires more investigation of public-sector sequential decision making applications, which provide unique challenges for machine learning researchers and can be socially beneficial. * Equal contribution. 2 We focus primarily on governance in the United States due to the expertise of authors, but most if not all of our examples have parallels in countries and regions around the world.

show abstract

Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond

Cited by 31 publications

References 80 publications

Using Sociolinguistic Variables to Reveal Changing Attitudes Towards Sexuality and Gender

Using Sociolinguistic Variables to Reveal Changing Attitudes Towards Sexuality and Gender

Interpreting the Robustness of Neural NLP Models to Textual Perturbations

Beyond Ads: Sequential Decision-Making Algorithms in Law and Public Policy

Contact Info

Product

Resources

About