Balancing the privacy-utility tradeoff is a crucial requirement of many practical machine learning systems that deal with sensitive customer data. A popular approach for privacy-preserving text analysis is noise injection, in which text data is first mapped into a continuous embedding space, perturbed by sampling a spherical noise from an appropriate distribution, and then projected back to the discrete vocabulary space. While this allows the perturbation to admit the required metric differential privacy, often the utility of downstream tasks modeled on this perturbed data is low because the spherical noise does not account for the variability in the density around different words in the embedding space. In particular, words in a sparse region are likely unchanged even when the noise scale is large. %Using the global sensitivity of the mechanism can potentially add too much noise to the words in the dense regions of the embedding space, causing a high utility loss, whereas using local sensitivity can leak information through the scale of the noise added. In this paper, we propose a text perturbation mechanism based on a carefully designed regularized variant of the Mahalanobis metric to overcome this problem. For any given noise scale, this metric adds an elliptical noise to account for the covariance structure in the embedding space. This heterogeneity in the noise scale along different directions helps ensure that the words in the sparse region have sufficient likelihood of replacement without sacrificing the overall utility. We provide a text-perturbation algorithm based on this metric and formally prove its privacy guarantees. Additionally, we empirically show that our mechanism improves the privacy statistics to achieve the same level of utility as compared to the state-of-the-art Laplace mechanism.

Deep Neural Networks, despite their great success in diverse domains, are provably sensitive to small perturbations on correctly classified examples and lead to erroneous predictions. Recently, it was proposed that this behavior can be combatted by optimizing the worst case loss function over all possible substitutions of training examples. However, this can be prone to weighing unlikely substitutions higher, limiting the accuracy gain. In this paper, we study adversarial robustness through randomized perturbations, which has two immediate advantages: (1) by ensuring that substitution likelihood is weighted by the proximity to the original word, we circumvent optimizing the worst case guarantees and achieve performance gains; and (2) the calibrated randomness imparts differentially-private model training, which additionally improves robustness against adversarial attacks on the model outputs. Our approach uses a novel density-based mechanism based on truncated Gumbel noise, which ensures training on substitutions of both rare and dense words in the vocabulary while maintaining semantic similarity for model robustness.

Membership Inference Attacks exploit the vulnerabilities of exposing models trained on customer data to queries by an adversary. In a recently proposed implementation of an auditing tool for measuring privacy leakage from sensitive datasets, more refined aggregates like the Log-Loss scores are exposed for simulating inference attacks as well as to assess the total privacy leakage based on the adversary's predictions. In this paper, we prove that this additional information enables the adversary to infer the membership of any number of datapoints with full accuracy in a single query, causing complete membership privacy breach. Our approach obviates any attack model training or access to side knowledge with the adversary. Moreover, our algorithms are agnostic to the model under attack and hence, enable perfect membership inference even for models that do not memorize or overfit. In particular, our observations provide insight into the extent of information leakage from statistical aggregates and how they can be exploited.

In (Yang et al. 2016), a hierarchical attention network (HAN) is created for document classification. The attention layer can be used to visualize text influential in classifying the document, thereby explaining the model's prediction. We successfully applied HAN to a sequential analysis task in the form of real-time monitoring of turn taking in conversations. However, we discovered instances where the attention weights were uniform at the stopping point (indicating all turns were equivalently influential to the classifier), preventing meaningful visualization for real-time human review or classifier improvement. We observed that attention weights for turns fluctuated as the conversations progressed, indicating turns had varying influence based on conversation state. Leveraging this observation, we develop a method to create more informative real-time visuals (as confirmed by human reviewers) in cases of uniform attention weights using the changes in turn importance as a conversation progresses over time.