recordLinkage

DSC 180B Project: Probabilistic Record Linkage

---

The machine learning pipeline can be broken down in 3 steps:

1. graph construction
2. node2vec embedding
3. classifier.

---

To run the program, refer to the config folder for the parameters that can be changed for this program, and use the python run.py command to train, test, and evualate the model.Refer to the run.py file for the appropriate command line inputs.

For visualizations and testing can be found in the notebooks directory.

The paper associated with this paper can be found here

---

Example Code

In order to generate the artificial dataset, you can use the command below:

python3 gen-data

In order to perform the graph construction, use the command:

python3 create-small-graphs

Lastly, in order to test any changes to the pipeline on example test, use the command:

python3 test-project

---

Configuration

1. refer to ./config/datasetGenConfig.json to change the hyperparameters associated with the artificial dataset generation.
2. refer to ./config/test_config.json and ./config/train_config.json to change the hyperparamters associated with generating the graphs and training the classifier.

(add information about the configurations)