SOTAVerified

Blocking

Entity resolution (also known as entity matching, record linkage, or duplicate detection) is the task of finding records that refer to the same real-world entity across different data sources (e.g., data files, books, websites, and databases). (Source: Wikipedia)

Blocking is a crucial step in any entity resolution pipeline because a pair-wise comparison of all records across two data sources is infeasible. Blocking applies a computationally cheap method to generate a smaller set of candidate record pairs reducing the workload of the matcher. During matching a more expensive pair-wise matcher generates a final set of matching record pairs.

Survey on blocking:

Papers

Showing 2130 of 524 papers

TitleStatusHype
AltDiffusion: A Multilingual Text-to-Image Diffusion ModelCode1
O^2-Recon: Completing 3D Reconstruction of Occluded Objects in the Scene with a Pre-trained 2D Diffusion ModelCode1
GraphSHA: Synthesizing Harder Samples for Class-Imbalanced Node ClassificationCode1
Path-Specific Counterfactual Fairness for Recommender SystemsCode1
Road Planning for Slums via Deep Reinforcement LearningCode1
Pre-trained Embeddings for Entity Resolution: An Experimental Analysis [Experiment, Analysis & Benchmark]Code1
Sparkly: A Simple yet Surprisingly Strong TF/IDF Blocker for Entity MatchingCode1
Tracker Meets Night: A Transformer Enhancer for UAV TrackingCode1
Self-Destructing Models: Increasing the Costs of Harmful Dual Uses of Foundation ModelsCode1
ABINet++: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text SpottingCode1
Show:102550
← PrevPage 3 of 53Next →

No leaderboard results yet.