Bibliographic record
Abstract
Data quality has become a pervasive challenge for organizations as they wrangle with large, heterogeneous datasets to extract value. Given the proliferation of sensitive and confidential information, it is crucial to consider data privacy concerns during the data cleaning process. For example, in medical database applications, varying levels of privacy are enforced across the attribute values. Attributes such as a patient’s country or city of residence may be less sensitive than the patient’s prescribed medication. Traditional data cleaning techniques assume the data is openly accessible, without considering the differing levels of information sensitivity. In this work, we take the first steps toward a data cleaning model that integrates privacy as part of the data cleaning process. We present a privacy-aware data cleaning framework that differentiates the information content among the attribute values during the data cleaning process to resolve data inconsistencies while minimizing the amount of information disclosed. Our data repair algorithm includes a set of data disclosure operations that considers the information content of the underlying attribute values, while maximizing data utility. Our evaluation using real datasets shows that our algorithm scales well, and achieves improved performance and comparable repair accuracy against existing data cleaning solutions.
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 | 0.004 | 0.017 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.005 | 0.005 |
| Science and technology studies | 0.002 | 0.001 |
| Scholarly communication | 0.008 | 0.008 |
| Open science | 0.004 | 0.007 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.066 | 0.051 |
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".