Frequency-based constraint relaxation for private query processing in cloud databases
Bibliographic record
Abstract
We introduce a new definition of privacy based on query frequencies, as well as a frequency-based constraint relaxation methodology for private queries. Private queries undergo processing so that users may obtain data from a database in such a way that the user's search intentions, i.e. the data which the user is interested in, will be protected against exposure. Most existing protocols for private querying rely on the following two constraints to achieve privacy: i) queries are encoded so that the database server can handle query processes but cannot actually decode queries; ii) the server is forced to check all data in the server when computing query results. Because of these constraints, even database servers cannot distinguish which data are selected from the database. However, this second constraint compels servers to spend O(n) computational cost for each query processed, where n is the number of data entries on the server. We introduce a weaker privacy condition which ensures that search intentions are hidden within a portion of the database, as opposed to ordinary private queries which hide search intentions among all data in the database, and we argue that this definition of privacy is sufficient to combat attacks based on query frequencies. Our relaxation methodology relaxes the second constraint above and allows private querying while only examining a portion of the data in most cases. Our methodology is also flexible and applies not only to exact match queries in one dimensional data but also to range queries in one dimensional data and exact match queries in two dimensional data.
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 | 0.007 | 0.024 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.004 | 0.010 |
| Open science | 0.003 | 0.005 |
| Research integrity | 0.001 | 0.005 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
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".