Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbations
Open accessThe DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Abstract
Accurately learning from user data while providing quantifiable privacy guarantees provides an opportunity to build better ML models while maintaining user trust. This paper presents a formal approach to carrying out privacy preserving text perturbation using the notion of d_χ-privacy designed to achieve geo-indistinguishability in location data. Our approach applies carefully calibrated noise to vector representation of words in a high dimension space as defined by word embedding models. We present a privacy proof that satisfies d_χ-privacy where the privacy parameter $\varepsilon$ provides guarantees with respect to a distance metric defined by the word embedding space. We demonstrate how $\varepsilon$ can be selected by analyzing plausible deniability statistics backed up by large scale analysis on GloVe and fastText embeddings. We conduct privacy audit experiments against $2$ baseline models and utility experiments on 3 datasets to demonstrate the tradeoff between privacy and utility for varying values of varepsilon on different task types. Our results demonstrate practical utility (< 2% utility loss for training binary classifiers) while providing better privacy guarantees than baseline models.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Enriching Word Vectors with Subword Information
Bojanowski, P., Grave, E., Joulin, A., & Mikolov, T. (2017) · Transactions of the Association for Computational Linguistics · doi:10.1162/tacl_a_00051 · Cited by 9,645 · cited 1×
“, how words cluster together and the distances between words and their neighbors – we ran several analytical experiments on two widely available word embedding models: GloVe [44] and fastText [9].”
MethodCalibrating Noise to Sensitivity in Private Data Analysis
Dwork, C., McSherry, F., Nissim, K., & Smith, A. (2006) · Lecture Notes in Computer Science · doi:10.1007/11681878_14 · Cited by 7,272
Membership Inference Attacks Against Machine Learning Models
Shokri, R., Stronati, M., Song, C., & Shmatikov, V. (2017) · 2017 IEEE Symposium on Security and Privacy (SP) · doi:10.1109/sp.2017.41 · Cited by 4,754 · cited 1×
“This could be either in the form of some shared side knowledge between the user and the analyst [33], or through an ML attack to learn which users frequently use a given set of words [51, 52].”
BackgroundRobust De-anonymization of Large Sparse Datasets
Narayanan, A., & Shmatikov, V. (2008) · 2008 IEEE Symposium on Security and Privacy (sp 2008) · doi:10.1109/sp.2008.33 · Cited by 2,372 · cited 1×
“Unintended disclosures, or accidental publications which lead to re-identification have been two common causes of recent privacy breaches [2, 7, 18, 39, 41, 55, 56].”
BackgroundPrivacy-preserving deep learning
Shokri, R., & Shmatikov, V. (2015) · Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/2810103.2813687 · Cited by 2,356 · cited 1×
“privacy guarantee (with ε =0representing absolute privacy, and ε =∞representing null privacy). Even thoughDP has been applied to domains suchas geolocation [4], social networks [40] and deep learning [1, 50], less attentionhasbeenpaidtoadaptingvariantsofDPtothecontextof Natural Language Processing (NLP) and the text d…”
MethodRAPPOR: Randomized Aggregatable Privacy-Preserving Ordinal Response
Erlingsson, Ú., Pihur, V., & Korolova, A. (2014) · Proceedings of the 2014 ACM SIGSAC Conference on Computer and Communications Security · doi:10.1145/2660267.2660348 · Cited by 1,838
Bidirectional LSTM Networks for Improved Phoneme Classification and Recognition
Graves, A., Fernández, S., & Schmidhuber, J. (2005) · Lecture Notes in Computer Science · doi:10.1007/11550907_126 · Cited by 925
Kasiviswanathan, S. P., Lee, H. K., Nissim, K., Raskhodnikova, S., & Smith, A. (2011) · SIAM Journal on Computing · doi:10.1137/090756090 · Cited by 898 · cited 8×
“Together, these requirements push us towards solutions where each data record is processed independently, similar to the setting in Local DP (LDP) [31].”
BackgroundMethodInfluentialC‐sanitized: A privacy model for document redaction and sanitization
Sánchez, D., & Batet, M. (2016) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.23363 · Cited by 80
Cited by
A Survey on Differential Privacy for Unstructured Data Content
Zhao, Y., & Chen, J. (2022) · ACM Computing Surveys · doi:10.1145/3490237 · Cited by 313
A Survey on Intelligent Internet of Things: Applications, Security, Privacy, and Future Directions
Aouedi, O., Vu, T.-H., Sacco, A., Nguyen, D. C., Piamrat, K., Marchetto, G., & Pham, Q.-V. (2025) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2024.3430368 · Cited by 277
How to DP-fy ML: A Practical Guide to Machine Learning with Differential Privacy
Ponomareva, N., Hazimeh, H., Kurakin, A., Xu, Z., Denison, C., McMahan, H. B., Vassilvitskii, S., Chien, S., & Thakurta, A. G. (2023) · Journal of Artificial Intelligence Research · doi:10.1613/jair.1.14649 · Cited by 166
“Recently Feyisetan et al. (2020) devise a dχ mechanism to modify a text string by taking each word, adding noise to its embedding, and then replacing the original word with the word closest to the noisy embedding.”
MethodThe Good and The Bad: Exploring Privacy Issues in Retrieval-Augmented Generation (RAG)
Zeng, S., Zhang, J., He, P., Liu, Y., Xing, Y., Xu, H., Ren, J., Chang, Y., Wang, S., Yin, D., et al. (2024) · Findings of the Association for Computational Linguistics ACL 2024 · doi:10.18653/v1/2024.findings-acl.267 · Cited by 149
LexGLUE: A Benchmark Dataset for Legal Language Understanding in English
Chalkidis, I., Jana, A., Hartung, D., Bommarito, M. J., Androutsopoulos, I., Katz, D. M., & Aletras, N. (2021) · SSRN Electronic Journal · doi:10.2139/ssrn.3936759 · Cited by 111
Metadata
- Authors
- Feyisetan, O., Balle, B., Drake, T., & Diethe, T.
- Venue
- Proceedings of the 13th International Conference on Web Search and Data Mining
- Publisher
- ACM
- Published
- 2020-01-20
- Pages
- 178-186
- OpenAlex
- W2980848703
- Cited by
- 115
- References
- 9
- FWCI
- 5.64
- Type
- conference paper
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbations. https://app.cochat.ai/papers/10.1145/3336191.3371856
- The paper
- Feyisetan, O., Balle, B., Drake, T., & Diethe, T. (2020). Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbations. Proceedings of the 13th International Conference on Web Search and Data Mining. https://doi.org/10.1145/3336191.3371856
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.