Hybrid Machine Learning and GenAI Approach for Data Loss Prevention
Bibliographic record
Abstract
Data leakage prevention (DLP) has become a critical requirement for organizations seeking to protect sensitive information in increasingly distributed and data-driven environments. Traditional rule-based DLP methods are brittle under paraphrase and format drift, while supervised approaches are hindered by the scarcity and cost of high-quality labeled data. This paper presents a hybrid pipeline that combines large language model-assisted labeling with a lightweight gradientboosted classifier for detecting confidential content in organizational textual communications. Using the publicly available U.S. Department of State email corpus, GPT-4 Turbo generated initial automated labels, which were validated on a manually reviewed sample before training. The resulting model, based on Term Frequency-Inverse Document Frequency (TF-IDF) features and XGBoost, achieved state-of-the-art performance compared to other ensemble and baseline classifiers, with $\mathbf{F}_{1}$-score of $\mathbf{7 6. 9 \%}$ and accuracy of 77%. These results demonstrate that large language models can effectively bootstrap high-quality annotations for sensitive content, enabling robust and interpretable DLP systems deployable in real-world contexts.
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 machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.003 | 0.003 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
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 source (direct Gemma or distilled Codex), 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".