Causal Estimation for Text Data with (Apparent) Overlap Violations

• URL: http://arxiv.org/abs/2210.00079v1
• Date: Fri, 30 Sep 2022 20:33:17 GMT
• Title: Causal Estimation for Text Data with (Apparent) Overlap Violations
• Authors: Lin Gui, Victor Veitch
• Abstract summary: We show how to handle causal identification and obtain robust causal estimation in the presence of apparent overlap violations. The idea is to use supervised representation learning to produce a data representation that preserves confounding information.
• Score: 16.94058221134916
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Consider the problem of estimating the causal effect of some attribute of a text document; for example: what effect does writing a polite vs. rude email have on response time? To estimate a causal effect from observational data, we need to adjust for confounding aspects of the text that affect both the treatment and outcome -- e.g., the topic or writing level of the text. These confounding aspects are unknown a priori, so it seems natural to adjust for the entirety of the text (e.g., using a transformer). However, causal identification and estimation procedures rely on the assumption of overlap: for all levels of the adjustment variables, there is randomness leftover so that every unit could have (not) received treatment. Since the treatment here is itself an attribute of the text, it is perfectly determined, and overlap is apparently violated. The purpose of this paper is to show how to handle causal identification and obtain robust causal estimation in the presence of apparent overlap violations. In brief, the idea is to use supervised representation learning to produce a data representation that preserves confounding information while eliminating information that is only predictive of the treatment. This representation then suffices for adjustment and can satisfy overlap. Adapting results on non-parametric estimation, we find that this procedure is robust to conditional outcome misestimation, yielding a low-bias estimator with valid uncertainty quantification under weak conditions. Empirical results show strong improvements in bias and uncertainty quantification relative to the natural baseline.

Related papers

• AMRFact: Enhancing Summarization Factuality Evaluation with AMR-Driven Negative Samples Generation [57.8363998797433]
  We propose AMRFact, a framework that generates perturbed summaries using Abstract Meaning Representations (AMRs) Our approach parses factually consistent summaries into AMR graphs and injects controlled factual inconsistencies to create negative examples, allowing for coherent factually inconsistent summaries to be generated with high error-type coverage.
  arXiv  Detail & Related papers  (2023-11-16T02:56:29Z)
• Interpretable Automatic Fine-grained Inconsistency Detection in Text Summarization [56.94741578760294]
  We propose the task of fine-grained inconsistency detection, the goal of which is to predict the fine-grained types of factual errors in a summary. Motivated by how humans inspect factual inconsistency in summaries, we propose an interpretable fine-grained inconsistency detection model, FineGrainFact.
  arXiv  Detail & Related papers  (2023-05-23T22:11:47Z)
• Improving Faithfulness of Abstractive Summarization by Controlling Confounding Effect of Irrelevant Sentences [38.919090721583075]
  We show that factual inconsistency can be caused by irrelevant parts of the input text, which act as confounders. We design a simple multi-task model to control such confounding by leveraging human-annotated relevant sentences when available. Our approach improves faithfulness scores by 20% over strong baselines on AnswerSumm citepfabbri 2021answersumm dataset.
  arXiv  Detail & Related papers  (2022-12-19T18:51:06Z)
• Revisiting text decomposition methods for NLI-based factuality scoring of summaries [9.044665059626958]
  We show that fine-grained decomposition is not always a winning strategy for factuality scoring. We also show that small changes to previously proposed entailment-based scoring methods can result in better performance.
  arXiv  Detail & Related papers  (2022-11-30T09:54:37Z)
• Neuro-Symbolic Entropy Regularization [78.16196949641079]
  In structured prediction, the goal is to jointly predict many output variables that together encode a structured object. One approach -- entropy regularization -- posits that decision boundaries should lie in low-probability regions. We propose a loss, neuro-symbolic entropy regularization, that encourages the model to confidently predict a valid object.
  arXiv  Detail & Related papers  (2022-01-25T06:23:10Z)
• Counterfactual Invariance to Spurious Correlations: Why and How to Pass Stress Tests [87.60900567941428]
  A spurious correlation' is the dependence of a model on some aspect of the input data that an analyst thinks shouldn't matter. In machine learning, these have a know-it-when-you-see-it character. We study stress testing using the tools of causal inference.
  arXiv  Detail & Related papers  (2021-05-31T14:39:38Z)
• Deconfounding Scores: Feature Representations for Causal Effect Estimation with Weak Overlap [140.98628848491146]
  We introduce deconfounding scores, which induce better overlap without biasing the target of estimation. We show that deconfounding scores satisfy a zero-covariance condition that is identifiable in observed data. In particular, we show that this technique could be an attractive alternative to standard regularizations.
  arXiv  Detail & Related papers  (2021-04-12T18:50:11Z)
• Double Perturbation: On the Robustness of Robustness and Counterfactual Bias Evaluation [109.06060143938052]
  We propose a "double perturbation" framework to uncover model weaknesses beyond the test dataset. We apply this framework to study two perturbation-based approaches that are used to analyze models' robustness and counterfactual bias in English.
  arXiv  Detail & Related papers  (2021-04-12T06:57:36Z)
• Statistically significant detection of semantic shifts using contextual word embeddings [7.439525715543974]
  We propose an approach to estimate semantic shifts by combining contextual word embeddings with permutation-based statistical tests. We demonstrate the performance of this approach in simulation, achieving consistently high precision by suppressing false positives. We additionally analyzed real-world data from SemEval-2020 Task 1 and the Liverpool FC subreddit corpus.
  arXiv  Detail & Related papers  (2021-04-08T13:58:54Z)
• Causal Effects of Linguistic Properties [41.65859219291606]
  We consider the problem of using observational data to estimate the causal effects of linguistic properties. We introduce TextCause, an algorithm for estimating causal effects of linguistic properties. We show that the proposed method outperforms related approaches when estimating the effect of Amazon review sentiment.
  arXiv  Detail & Related papers  (2020-10-24T15:43:37Z)
• Semantic Neighborhood-Aware Deep Facial Expression Recognition [14.219890078312536]
  A novel method is proposed to formulate semantic perturbation and select unreliable samples during training. Experiments show the effectiveness of the proposed method and state-of-the-art results are reported.
  arXiv  Detail & Related papers  (2020-04-27T11:48:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.