- Create a dataset using ORE by running VectorPrinting experiment with respect to the selected domain of interest and schema matchers.
1.1 An example dataset is available for download: Beta Dataset - Clone the ADnEV repository
- Update Config with your configuration details.
- Run mainSaver to train and test your dataset using a 5-fold cross validation.
1.1 You can also run a pre-trained model using mainLoader. - The results will appear in the results folder, there you will find a notebook to help you analyze the results.
- Your models will appear in the models folder, there you will find a notebook to help you visualize the models.
ADnEV: Cross-Domain Schema Matching using Deep Similarity Matrix Adjustment and Evaluation. Roee Shraga, Avigdor Gal, Haggai Roitman, PVLDB, 9(13):1401-1415, 2020.
BibTeX:
@article{shraga2020,
title={ADnEV: Cross-Domain Schema Matching using Deep Similarity Matrix Adjustment and Evaluation},
author={Shraga, Roee and Gal, Avigdor and Roitman, Haggai},
journal={Proceedings of the VLDB Endowment},
volume={13},
number={9},
pages={1401--1415},
year={2020},
publisher={VLDB Endowment}
}
ADnEV was developed at the Technion - Israel Institute of Technology by Roee Shraga under the supervision of Prof. Avigdor Gal in collaboration with Haggai Roitman from IBM Research - AI.