Advanced techniques for high performance query optimization in database systems
Bibliographic record
Abstract
Current query optimizers employ many sources of information about the database to optimize queries. Statistics and integrity constraints—primary key and foreign key constraints in particular—have long played a role in the optimizer. Strategies that exploit constraints have been seen to offer good improvement for query evaluation. A problem, however, is that often the “constraints” that would be useful for optimization for a given database are not explicitly available for the optimizer. In this dissertation we investigate how to discover useful constraint-like characterizations of large data sets via data mining. Specifically, we propose join holes as a new way for characterizing data values for multiple attributes. An efficient and scalable algorithm is presented for identifying all maximal join holes over large data sets. We explore the applications of join holes for query optimization in two areas: query rewrite and join size estimation. We propose SIEQE, a new statistic built over non base tables by means of join holes for cardinality prediction. The proposed techniques extend the functionalities of current query optimizers and provide better estimation of cardinality for join queries in an efficient way. Database systems nowadays make use of materialized views for the efficiency of query evaluation. Determining which views to materialize, however, is a non-trivial task. We address the view selection problem in this dissertation by presenting an algorithm to automatically recommend vew to materialize for given workload queries. Our work in this area has been implemented in latest DB2 product which demonstrated significant improvement. Skyline is a proposed query operator that could be useful in expressing preference queries. Computing skyline queries expressed in standard SQL statements would be cumbersome and expensive with current database optimization. We address this by developing sort-filter-skyline, an efficient algorithm that is well-behaved in a relational context, and is easily accommodated by the query optimizer.
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 | 0.016 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.003 | 0.003 |
| Bibliometrics | 0.003 | 0.008 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.006 | 0.009 |
| Open science | 0.004 | 0.005 |
| Research integrity | 0.002 | 0.007 |
| Insufficient payload (model declined to judge) | 0.007 | 0.004 |
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".