iCAT+: An Interactive Customizable Anonymization Tool Using Automated Translation Through Deep Learning
Bibliographic record
Abstract
Data anonymization is a viable solution for data owners to mitigate their privacy concerns. However, existing data anonymization tools are inflexible to support various privacy and utility requirements of both data owners and data users. In most cases, this limitation is due to a lack of understanding of those requirements as well as the non-customizability of the existing tools. To address this limitation, we propose <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">iCAT+</i>, which is an interactive and customizable anonymization approach. More specifically, we first automate the interpretation of data owners’ and data users’ textual requirements by deploying a Convolutional Neural Network (CNN) model for Natural Language Processing (NLP). Second, we introduce the concept of the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">anonymization space</i> to model possible combinations of per-attribute anonymization primitives based on the level of privacy and utility that each primitive provides. Third, we design an ontology model that maps the translated requirements into their appropriate anonymization primitives in the defined anonymization space corresponding to the plain data. Fourth, we evaluate the efficiency and effectiveness of <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">iCAT+</i> based on both real and synthetic network data. Finally, we assess its usability through a real user study involving participants from industry and research laboratories. Our experiments show the effectiveness and efficiency of our solution (e.g., requirement translation accuracy of 99% at the data owner side and 98% at the data user side, with a computational time of around one minute for the Google cluster dataset).
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.002 |
| Open science | 0.003 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".