From Probabilistic to Fuzzy Matching Record Linkage: A promising Transition
Bibliographic record
Abstract
ObjectiveProbabilistic Record Linkage (PRL) heavily relies on manual intervention for gray area resolutions. This makes PRL extremely time and resource intensive. No matter how scientifically sound and robust PRL method is, it didn’t meet close-to-real-time data availability requirements at ICES. ApproachDuring a thorough evaluation and comparison process of three software and two methods of Record Linkage, ICES initiated a semi-design of experiment to select an optimal record linkage approach. For this experiment, a large Ontario data on 12 million individuals with required linkage variables plus valid Ontario health card numbers was selected. While the availability of the health card number enabled assessment of the accuracy of different approaches, the analysts were blinded to the correct health card numbers during the process. If manual intervention was required, it was repeated by two analysts to allow capturing human error. ResultsPRL-based software needed the most personnel time to complete the process. Human errors were identified during the manual intervention due to subjective decisions by the analysts. Fuzzy Matching approach eliminated manual intervention but achieved comparable linkage rate to PRL while maintaining the same accuracy. The Fuzzy Matching software costs were higher; however, the data timeliness was significantly improved, and the clerical review costs and human error were eliminated. ConclusionsThe Modernization of Record Linkage (MORL) project was a successful demonstration of the advantages of Fuzzy Matching over PRL method. However, the implementation of new approach at the organization level was a challenging change management.
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 | 0.009 | 0.005 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.006 | 0.007 |
| Open science | 0.004 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| 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".