Implementing Informative-Based Active Learning in Biomedical Record Linkage for the Splink Package in Python.
Marko MileticMurat SariyarPublished in: ICIMTH (2023)
Keyphrases
- record linkage
- active learning
- data cleaning
- privacy preserving
- duplicate detection
- entity resolution
- multiple databases
- programming language
- active learner
- open source
- learning algorithm
- group membership
- census data
- approximate matching
- random sampling
- machine learning
- training set
- linked data
- information extraction
- cost sensitive
- disclosure risk
- open source software
- information content
- labeled data
- semi supervised
- artificial intelligence
- text mining