pyjedai.comparison_cleaning

pyjedai.comparison_cleaning#

Functions

get_meta_blocking_approach(acronym, w_scheme)

Return method by acronym

Classes

AbstractComparisonCleaning()

Abstract class for Block cleaning

AbstractMetablocking()

Restructure a redundancy-positive block collection into a new one that contains substantially lower number of redundant and superfluous comparisons, while maintaining the original number of matching ones

BLAST([weighting_scheme])

Meta-blocking method that retains the comparisons that correspond to edges in the blocking graph that are exceed 1/4 of the sum of the maximum edge weights in the two adjacent node neighborhoods.

CardinalityEdgePruning([weighting_scheme])

A Meta-blocking method that retains the comparisons that correspond to the top-K weighted edges in the blocking graph.

CardinalityNodePruning([weighting_scheme])

A Meta-blocking method that retains for every entity, the comparisons that correspond to its top-k weighted edges in the blocking graph."

ComparisonPropagation()

Comparison Propagation

GlobalProgressiveSortedNeighborhood([...])

Global Progressive Sorted Neighborhood

LocalProgressiveSortedNeighborhood([...])

Local Progressive Sorted Neighborhood

ProgressiveCardinalityEdgePruning([...])

ProgressiveCardinalityNodePruning([...])

ProgressiveEntityScheduling([...])

Progressive Entity Scheduling

ProgressiveSortedNeighborhood([...])

Progressive Sorted Neighborhood

ReciprocalCardinalityNodePruning([...])

A Meta-blocking method that retains the comparisons that correspond to edges in the blocking graph that are among the top-k weighted ones for both adjacent entities/nodes.

ReciprocalWeightedNodePruning([weighting_scheme])

Meta-blocking method that retains the comparisons that correspond to edges in the blocking graph that are exceed the average edge weight in both adjacent node neighborhoods.

WeightedEdgePruning([weighting_scheme])

A Meta-blocking method that retains all comparisons that have a weight higher than the average edge weight in the blocking graph.

WeightedNodePruning([weighting_scheme])

A Meta-blocking method that retains for every entity, the comparisons that correspond to edges in the blocking graph that are exceed the average edge weight in the respective node neighborhood.