The Use of Smart Tokens in Cleaning Integrated Warehouse Data
Bibliographic record
Abstract
Identifying integrated records that represent the same real-world object in numerous ways is just one form of data disparity (dirt) to be resolved in a data warehouse. Data cleaning is a complex process, which uses multidisciplinary techniques to resolve conflicts in data drawn from different data sources. There is a need for initial cleaning at the time a data warehouse is built, and incremental cleaning whenever new records are brought into the data warehouse during refreshing. Existing work on data cleaning have used pre-specified record match thresholds and multiple scanning of records to determine matching records in integrated data. Little attention has been paid to incremental matching of records. Determining optimal record match score threshold in a domain is hard. Also, direct long record string comparison is highly inefficient and intolerant to typing errors. Thus, this article proposes two algorithms, the first of which uses smart tokens defined from integrated records to match and identify duplicate records during initial warehouse cleaning. The second algorithm uses these tokens for fast, incremental cleaning during warehouse refreshing. Every attribute value forms either a special token like birth date or an ordinary token, which can be alphabetic, numeric, or alphanumeric. Rules are applied for forming tokens belonging to each of these four classes. These tokens are sorted and used for record match. The tokens also form very good warehouse identifiers for future faster incremental warehouse cleaning. This approach eliminates the need for match threshold and multiple passes at data. Experiments show that using tokens for record comparison produces a far better result than using the entire or greater part of a record.
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 | 0.003 | 0.012 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.003 | 0.006 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.006 |
| Open science | 0.004 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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".