anti-hub recall instrument
| Definition | Recall at k restricted to the rows that are rarely retrieved, the observer's outliers, reported as the minimum over strata so that a mean cannot hide the tail. Chapter 10. |
|---|---|
| Example | Seven strata of rarely retrieved rows scored between 0.62 and 0.69, so the minimum, 0.663, failed the bar of 0.90. |
| Book | Data Mining as Observation, draft 0.2, commit f3914f0; entry id anti-hub-recall, kind instrument. |
| Status | no ledger row names this entry. Corrections: none recorded. |
| Defining equation | none |
| Assumptions and scope |
|
| Prior art | none recorded |
| Evidence | turboquant-pro/docs/HUBNESS_PRIMER.md:86-131, lean/DataMiningAsObservation/RecallAtK.lean, lean/DataMiningAsObservation/Hubness.lean |
| Reviewed | not yet reviewed; generated 2026-09-10 from records at the commits on the provenance page. |
Equation
none
Conditions
- Recall at k restricted to the rows that are rarely retrieved, the observer’s outliers, reported as the minimum over strata so that a mean cannot hide the tail. It lies in the unit interval and is one exactly when every anti-hub’s true neighbours are returned.
- Compressed indexes fail on anti-hubs first while aggregate recall barely moves, and the remedy that promotes isolated rows is the remedy that compression damages first, which is why the gate is a build gate and not a report line.
Conditions are curated in entries.toml rather than read
from a record.
Ledger
none
First stated
Chapter 10 section 10.2 of Data Mining as Observation, with
the program’s gate in
turboquant-pro/docs/HUBNESS_PRIMER.md:86-131 and the strata
result in
turboquant-pro/docs/RESULTS_strata_phase23_gates.md.
Measurements
| Where the book states it | Numbers, as the book’s sources table records them | Source |
|---|---|---|
| chapter 11 section 11.6 | anti-hub recall, p05, hub-rank correlation, hub-set overlap, the build gate | turboquant-pro/docs/HUBNESS_PRIMER.md:86-131 |
Failures and corrections
none
Invariance envelope
none declared
Machine checked
lean/DataMiningAsObservation/RecallAtK.lean,
theorems recallAtK_mem_unit,
recallAtK_eq_one_iff, aggregate_le_of_failing,
aggregate_example, at observation-data-mining f3914f0; what
the check covers is stated in the book’s appendix
C.
lean/DataMiningAsObservation/Hubness.lean,
theorems sum_count, sum_count_eq,
count_congr, antiHub_iff, at
observation-data-mining f3914f0; what the check covers is stated in the
book’s appendix
C.
Used in
Data Mining as Observation chapters 10, 11, 12.
Related
anti-hub; recall at k; min-over-strata; hub; stratification.
See also
Book equations stated beside the entry’s terms, not defining it: 10.7, 11.3, 0.35.
Ledger rows that cite the entry’s records without naming it: NEG-14.
Sources-table rows that share a record with the entry without naming it: chapter 10 section 10.2, chapter 10 section 10.4, chapter 10 section 10.5, chapter 12 section 12.4.
Status
Generated 2026-09-10 by encyclopedia/generate.py; book
at observation-data-mining f3914f0; the commit of every record is listed
in the encyclopedia’s provenance.