This folder contains examples on record linkage with the Python Record Linkage Toolkit. The examples do have a BSD 3-Clause "New" or "Revised" License. Contributions are appreciated.
Example of deterministic record linkage to find duplicated records in a dataset. In this example, the model isn't trained with train data.
Example of deterministic record linkage to find links between two datasets. In this example, the model isn't trained with train data.
An implementation of the Fellegi and Sunter (1969) classification model in a supervised way.
An implementation of the Fellegi and Sunter (1969) classification model in an unsupervised way. The training of model parameters is done with the Expectation-Conditional Maximisation algorithm.
This example shows how Neural Networks can be used to classify record pairs. The Neural Network is implemented in Keras.