Skip to main content
SpringerLink
Log in

Causal Inference and Natural Language Processing

  • Chapter
  • First Online:
Machine Learning for Causal Inference

  • 758 Accesses

Abstract

This chapter explores the intersection of two research fields: causal inference and natural language processing (NLP). We aim to answer two fundamental questions: (1) how can NLP aid in causal inference when working with textual data, and (2) how can causal inference theory enhance the robustness and interpretability of NLP models? We present the latest developments and challenges in each area. Firstly, we discuss the difficulties associated with performing causal inference with textual data, which stems from the unstructured and high-dimensional nature of the text. We demonstrate how NLP models can extract high-level semantic variables and how textual data can assume various roles in the causal graph based on Pearl’s causal theory. Secondly, while NLP models have achieved remarkable success across different tasks, we highlight concerns about their reliability and robustness. NLP models are prone to learning spurious correlations, which are non-causal but correlated relationships. Thirdly, we provide an extensive overview of causality-driven models for NLP, examining various methods of integrating causality, including intervention-level and counterfactual-level debiasing techniques. Finally, we explore how causal interpretations can improve the interpretability of deep neural models in NLP, enabling a more profound understanding of the models.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 129.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Hardcover Book
USD 169.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    https://gluebenchmark.com/leaderboard, https://super.gluebenchmark.com/leaderboard

References

  1. R. Aralikatte et al., Focus attention: promoting faithfulness and diversity in summarization, in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) (2021), pp. 6078–6095

    Google Scholar 

  2. M. Arjovsky et al., Invariant risk minimization (2019). arXivabs/1907.02893

    Google Scholar 

  3. D. Bahdanau, K. Cho, Y. Bengio, Neural machine translation by jointly learning to align and translate, in 3rd International Conference on Learning Representations, ICLR (2015)

    Google Scholar 

  4. E. Bareinboim, J. Pearl, Controlling selection bias in causal inference, in Proceedings of the Fifteenth International Conference on Artificial Intelligence and Statistics, PMLR. vol. 22 (2012), pp. 100–108

    Google Scholar 

  5. E. Bareinboim et al., On pearl’s hierarchy and the foundations of causal inference, in Probabilistic and Causal Inference (2022)

    Google Scholar 

  6. Y. Belinkov, S. Gehrmann, E. Pavlick, Interpretability and analysis in neural NLP, in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: Tutorial Abstracts (2020), pp. 1–5

    Google Scholar 

  7. Y. Belinkov, J. Glass, Analysis methods in neural language processing: a survey, Trans. Assoc. Comput. Linguist. 7, 49–72 (2019)

    Article  Google Scholar 

  8. R. Bommasani, C. Cardie, Intrinsic evaluation of summarization datasets, in Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, EMNLP (2020), pp. 8075–8096

    Google Scholar 

  9. W. Chen et al., De-confounded variational encoder-decoder for logical table-to-text generation, in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL-IJCNN (2021), pp. 5532–5542

    Google Scholar 

  10. W. Chen et al., Dependent multi-task learning with causal intervention for image captioning, in Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence, IJCAI-21, eds. by Z.-H. Zhou. Main Track. International Joint Conferences on Artificial Intelligence Organization (2021), pp. 2263–2270. https://doi.org/10.24963/ijcai.2021/312

  11. W. Chen et al., Exploring logically dependent multi-task learning with causal inference, in Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) (2020), pp. 2213–2225

    Google Scholar 

  12. S. Choudhary, N. Chatterjee, S.K. Saha, Interpretation of black box NLP models: a survey (2022). arXiv preprint arXiv:2203.17081

    Google Scholar 

  13. M. Cornia et al., Meshed-memory transformer for image captioning, in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR (2020), pp. 10575–10584

    Google Scholar 

  14. J. Devlin et al., BERT: pre-training of deep bidirectional transformers for language understanding, in: Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2019), pp. 4171–4186

    Google Scholar 

  15. N. Egami et al., How to make causal inferences using texts (2018). arXiv abs/1802.02163

    Google Scholar 

  16. N. Egami et al., How to make causal inferences using texts. Sci. Adv. 8(42) (2022). eabg2652. https://www.science.org/doi/pdf/10.1126/sciadv.abg2652

  17. Y. Elazar et al., Amnesic probing: behavioral explanation with amnesic counterfactuals. Trans. Assoc. Comput. Linguist. 9, 160–175 (2021)

    Article  Google Scholar 

  18. A. Feder et al., Causal inference in natural language processing: estimation, prediction, interpretation and beyond (2021). arXiv abs/2109.00725

    Google Scholar 

  19. A. Feder et al., CausaLM: causal model explanation through counterfactual language models. Comput. Linguist. 47(2), 333–386 (2021)

    Google Scholar 

  20. Y. Feng et al., Modeling fluency and faithfulness for diverse neural machine translation. Proc. AAAI Conf. Artif. Intell. 34(01), 59–66 (2020)

    Google Scholar 

  21. M. Finlayson et al., Causal analysis of syntactic agreement mechanisms in neural language models, in Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL-IJCNLP 2021 (Association for Computational Linguistics (ACL), 2021), pp. 1828–1843

    Google Scholar 

  22. C. Fong, J. Grimmer, Discovery of treatments from text corpora, in Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, ACL (2016), pp. 1600–1609

    Google Scholar 

  23. S. Garg et al., Counterfactual fairness in text classification through robustness, in Proceedings of the 2019 AAAI/ACM Conference on AI, Ethics, and Society, AIES (2019), pp. 219–226

    Google Scholar 

  24. M. Gill, A.B. Hall, How judicial identity changes the text of legal rulings, in Political Methods: Quantitative Methods eJournal (2015)

    Google Scholar 

  25. S. Gururangan et al., Annotation artifacts in natural language inference data, in 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2018), pp. 107–112

    Google Scholar 

  26. I. Habernal et al., The argument reasoning comprehension task: identification and reconstruction of implicit warrants, in Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2018), pp. 1930–1940

    Google Scholar 

  27. D. Hendrycks, K. Lee, M. Mazeika, Using pre-training can improve model robustness and uncertainty, in Proceedings of the 36th International Conference on Machine Learning, ICML, vol. 97. Proceedings of Machine Learning Research (2019), pp. 2712–2721

    Google Scholar 

  28. D. Hendrycks et al., Pretrained transformers improve out-of-distribution robustness, in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL (2020), pp. 2744–2751

    Google Scholar 

  29. D. Hovy, S. Prabhumoye, Five sources of bias in natural language processing. Lang. Linguist. Compass 15(8), e12432 (2021)

    Google Scholar 

  30. D. Hovy, A. Søgaard, Tagging performance correlates with author age, in Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing, ACL-IJCNLP (2015), pp. 483–488

    Google Scholar 

  31. G. Hripcsak et al., Causal inference from observational healthcare data: implications, impacts and innovations, in American Medical Informatics Association Annual Symposium, AMIA (2020)

    Google Scholar 

  32. Z. Hu, L.E. Li, A causal lens for controllable text generation. Adv. Neural Inf. Process. Syst. 34, 24941–24955 (2021)

    Google Scholar 

  33. S. Iyer, N. Dandekar, K. Csernai et al., First quora dataset release: question pairs (2017). data.quora.com

    Google Scholar 

  34. Z. Ji et al., Survey of hallucination in natural language generation, in ACM Computing Surveys (2022)

    Google Scholar 

  35. D. Kaushik, E.H. Hovy, Z.C. Lipton, Learning the difference that makes a difference with counterfactually-augmented data, in 8th International Conference on Learning Representations, ICLR (2020)

    Google Scholar 

  36. K. Keith, D. Rice, B. O’Connor, Text as causal mediators: research design for causal estimates of differential treatment of social groups via language aspects, in Proceedings of the First Workshop on Causal Inference and NLP (2021), pp. 21–32

    Google Scholar 

  37. K.A. Keith, D. Jensen, B. O’Connor. Text and causal inference: a review of using text to remove confounding from causal estimates, in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, ACL (2020), pp. 5332–5344

    Google Scholar 

  38. V. Landeiro, A. Culotta, Robust text classification under confounding shift. J. Artif. Intell. Res. 63, 391–419 (2018)

    Article  MathSciNet  Google Scholar 

  39. V. Landeiro, T. Tran, A. Culotta, Discovering and controlling for latent confounds in text classification using adversarial domain adaptation, in Proceedings of the 2019 SIAM International Conference on Data Mining, SDM (2019), pp. 298–305

    Google Scholar 

  40. H. Li et al., Ensure the correctness of the summary: incorporate entailment knowledge into abstractive sentence summarization, in Proceedings of the 27th International Conference on Computational Linguistics (2018), pp. 1430–1441

    Google Scholar 

  41. M. Li et al., Learning to imagine: integrating counterfactual thinking in neural discrete reasoning, in Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (2022), pp. 57–69

    Google Scholar 

  42. A. Lin et al. One-stage deep instrumental variable method for causal inference from observational data, in 2019 IEEE International Conference on Data Mining, ICDM (2019), pp. 419–428

    Google Scholar 

  43. B. Liu et al., Show, deconfound and tell: image captioning with causal inference, in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (2022), pp. 18041–18050

    Google Scholar 

  44. Y. Liu et al., RoBERTa: a robustly optimized bert pretraining approach (2019). arXiv abs/1907.11692

    Google Scholar 

  45. C. Louizos et al., Causal effect inference with deep latent-variable models, in Annual Conference on Neural Information Processing Systems 2017, NeurIPS (2017), pp. 6446–6456

    Google Scholar 

  46. A. Madsen, S. Reddy, S. Chandar, Post-hoc interpretability for neural NLP: a survey (2021). arXiv preprint arXiv:2108.04840

    Google Scholar 

  47. T. McCoy, E. Pavlick, T. Linzen, Right for the wrong reasons: diagnosing syntactic heuristics in natural language inference, in Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL (2019), pp. 3428–3448

    Google Scholar 

  48. W. Miao, Z. Geng, E.J. Tchetgen Tchetgen, Identifying causal effects with proxy variables of an unmeasured confounder. Biometrika 105(4), 987–993 (2018)

    Article  MathSciNet  MATH  Google Scholar 

  49. R. Moraffah et al., Causal interpretability for machine learning-problems, methods and evaluation. ACM SIGKDD Explorations Newslett. 22(1), 18–33 (2020)

    Article  Google Scholar 

  50. A. Naik et al., Stress test evaluation for natural language inference, in Proceedings of the 27th International Conference on Computational Linguistics, COLING (2018), pp. 2340–2353

    Google Scholar 

  51. R. allapati et al. Abstractive text summarization using sequence-to-sequence RNNs and beyond, in Proceedings of The 20th SIGNLL Conference on Computational Natural Language Learning (2016), pp. 280–290

    Google Scholar 

  52. A. Nichols, Causal inference with observational data. Stata J. 7(4), 507–541 (2007)

    Article  Google Scholar 

  53. T. Niven, H.-Y. Kao, Probing neural network comprehension of natural language arguments, in Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL (2019), pp. 4658–4664

    Google Scholar 

  54. Y. Pan et al., X-linear attention networks for image captioning, in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR (2020), pp. 10968–10977

    Google Scholar 

  55. J. Pearl, Causality, 2nd ed. (Cambridge University Press, Cambridge, 2009)

    Book  MATH  Google Scholar 

  56. J. Pearl, Theoretical impediments to machine learning with seven sparks from the causal revolution (2018). arXiv preprint arXiv:1801.04016

    Google Scholar 

  57. J. Pearl, D. Mackenzie, The Book of Why: The New Science of Cause and Effect, 1st edn. (Basic Books, Inc., New York, 2018)

    MATH  Google Scholar 

  58. A. Perez-Suay, G. Camps-Valls, Causal inference in geoscience and remote sensing from observational data. IEEE Trans. Geosci. Remote. Sens. 57(3), 1502–1513 (2019)

    Article  Google Scholar 

  59. M. Peyrard et al., Invariant language modeling, in EMNLP 2022 (2021)

    Google Scholar 

  60. R. Pryzant, Y. Chung, D. Jurafsky, Predicting sales from the language of product descriptions, in Proceedings of the SIGIR 2017 Workshop On eCommerce Co-located with the 40th International ACM SI-GIR Conference on Research and Development in Information Retrieval, eCOM@SIGIR (2017)

    Google Scholar 

  61. R. Pryzant et al., Causal effects of linguistic properties, in NAACL-HLT (2021)

    Google Scholar 

  62. R. Pryzant et al., Causal effects of linguistic properties, in Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2021), pp. 4095–4109

    Google Scholar 

  63. R. Pryzant et al., Deconfounded lexicon induction for interpretable social science, in Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2018), pp. 1615–1625

    Google Scholar 

  64. A. Radford et al., Language models are unsupervised multitask learners. OpenAI Blog 1(8), 9 (2019)

    Google Scholar 

  65. M.E. Roberts, B.M. Stewart, R.A. Nielsen, Adjusting for Confounding with Text Matching. Am. J. Polit. Sci. 64, 887–903 (2020)

    Article  Google Scholar 

  66. J.M. Rohrer, Thinking clearly about correlations and causation: Graphical causal models for observational data. Adv. Methods Practices Psychol. Sci. 1(1), 27–42 (2018)

    Article  MathSciNet  Google Scholar 

  67. A. Ross, A. Marasović, M.E. Peters, Explaining NLP models via minimal contrastive editing (MiCE), in Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 (2021), pp. 3840–3852

    Google Scholar 

  68. D.B. Rubin, Estimating causal effects of treatments in randomized and nonrandomized studies. J. Educ. Psychol. 66(5), 688 (1974)

    Google Scholar 

  69. C. Rudin, Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 1(5), 206–215 (2019)

    Article  Google Scholar 

  70. B. Schölkopf, Causality for machine learning, in Probabilistic and Causal Inference: The Works of Judea Pearl (2022), pp. 765–804

    Google Scholar 

  71. B. Schölkopf et al., Toward causal representation learning. Proc. IEEE 109(5), 612–634 (2021)

    Article  Google Scholar 

  72. R. Shekhar et al., FOIL it! Find One mismatch between Image and Language caption, in Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, ACL (2017), pp. 255–265

    Google Scholar 

  73. C. Shorten, T.M. Khoshgoftaar, B. Furht, Text data augmentation for deep learning. J. Big Data 8, 1–34 (2021)

    Article  Google Scholar 

  74. H.A. Simon, Spurious correlation: a causal interpretation. J. Am. Statis. Assoc. 49(267), 467–479 (1954)

    MATH  Google Scholar 

  75. D. Sridhar, D.M. Blei, Causal inference from text: a commentary. Sci. Adv. 8(42), eade6585 (2022)

    Google Scholar 

  76. D. Sridhar, L. Getoor, Estimating causal effects of tone in online debates, in Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI (2019), pp. 1872–1878

    Google Scholar 

  77. N. Tandon et al., WIQA: a dataset for “What if…” reasoning over procedural text, in Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) (2019), pp. 6076–6085

    Google Scholar 

  78. B. Tian et al., Debiasing NLU models via causal intervention and counterfactual reasoning. Proc. AAAI Conf. Artif. Intell. 36(10), 11376–11384 (2022)

    Google Scholar 

  79. R. Tian et al., Sticking to the facts: confident decoding for faithful data-to-text generation (2019). arXiv preprint arXiv:1910.08684

    Google Scholar 

  80. L. Tu et al., An empirical study on robustness to spurious correlations using pre-trained language models. Trans. Assoc. Comput. Linguist. 8, 621–633 (2020)

    Article  Google Scholar 

  81. V. Veitch, D. Sridhar, D.M. Blei, Adapting text embeddings for causal inference, in Proceedings of the Thirty-Sixth Conference on Uncertainty in Artificial Intelligence, UAI, vol. 124. Proceedings of Machine Learning Research (2020), pp. 919–928

    Google Scholar 

  82. J. Vig et al., Causal mediation analysis for interpreting neural NLP: the case of gender bias (2020). arXiv preprint arXiv:2004.12265

    Google Scholar 

  83. O. Vinyals, Q.V. Le, A neural conversational model, in ICML Deep Learning Workshop (2015)

    Google Scholar 

  84. E. Wallace, M. Gardner, S. Singh, Interpreting predictions of NLP models, in Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: Tutorial Abstracts (2020), pp. 20–23

    Google Scholar 

  85. A. Wang et al., GLUE: a multi-task benchmark and analysis platform for natural language understanding, in 7th International Conference on Learning Representations, ICLR (2019)

    Google Scholar 

  86. T. Wang et al., Visual Commonsense R-CNN, in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR (2020), pp. 10757–10767

    Google Scholar 

  87. X. Wang, H. Wang, D. Yang, Measure and improve robustness in NLP models: a survey, in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (2022), pp. 4569–4586

    Google Scholar 

  88. Z. Wang, A. Culotta, Identifying spurious correlations for robust text classification, in Findings of the Association for Computational Linguistics: EMNLP (2020), pp. 3431–3440

    Google Scholar 

  89. Z. Wang, A. Culotta, Robustness to spurious correlations in text classification via automatically generated counterfactuals, in Thirty-Fifth AAAI Conference on Artificial Intelligence, AAAI (2021), pp. 14024–14031

    Google Scholar 

  90. J. Wen et al., AutoCAD: automatically generating counterfactuals for mitigating shortcut learning (2022). arXiv preprint arXiv:2211.16202

    Google Scholar 

  91. T. Wu et al., Polyjuice: generating counterfactuals for explaining, evaluating, and improving models, in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL-IJCNN (2021), pp. 6707–6723

    Google Scholar 

  92. K. Xu et al., Show, attend and tell: neural image caption generation with visual attention, in Proceedings of the 32nd International Conference on Machine Learning, ICML, vol. 37. JMLR Workshop and Conference Proceedings (2015), pp. 2048–2057

    Google Scholar 

  93. K. Xu et al., Show, attend and tell: neural image caption generation with visual attention, in International Conference on Machine Learning. PMLR (2015), pp. 2048–2057

    Google Scholar 

  94. X. Yang, H. Zhang, J. Cai, Deconfounded image captioning: a causal retrospect, in IEEE Transactions on Pattern Analysis and Machine Intelligence (2021)

    Google Scholar 

  95. Y. Zhang, J. Baldridge, L. He, PAWS: paraphrase adversaries from word scrambling, in Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT (2019), pp. 1298–1308

    Google Scholar 

  96. M. Zhou et al., Progress in neural NLP: modeling, learning, and reasoning. Engineering 6(3), 275–290 (2020)

    Article  MathSciNet  Google Scholar 

  97. F. Zhu et al., TAT-QA: a question answering benchmark on a hybrid of tabular and textual content in finance, in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) (2021), pp. 3277–3287

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Sun Yat-sen University, Zhuhai, China

    Wenqing Chen

  2. Ant Group, Hangzhou, China

    Zhixuan Chu

Authors
  1. Wenqing Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zhixuan Chu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Wenqing Chen .

Editor information

Editors and Affiliations

  1. University of Virginia, Charlottesville, VA, USA

    Sheng Li

  2. Ant Group, Hangzhou, China

    Zhixuan Chu

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Check for updates. Verify currency and authenticity via CrossMark

Cite this chapter

Chen, W., Chu, Z. (2023). Causal Inference and Natural Language Processing. In: Li, S., Chu, Z. (eds) Machine Learning for Causal Inference. Springer, Cham. https://doi.org/10.1007/978-3-031-35051-1_9

Download citation

Publish with us

Policies and ethics

Search

Navigation