Code related to our ISWC 2023 submission
- data: The datasets used in this paper, including the original and our splits.
- scripts: The Transformer, ConvSeq2Seq, BART, and T5 models, as well as the scripts to train and evaluate the models, and the script to produce the new splits
- demo.ipynb: The demonstration notebook that includes training and evaluation of a model as well as using the split algorithm.
If you use our work, please cite the following paper:
Samuel Reyd et Amal Zouaq, “Assessing the generalization capabilities of neural machine translation models for SPARQL query generation,” in The 22nd International Semantic Web Conference (ISWC) (accepted), 2023