Datasets:
Link to paper in readme
Browse files
README.md
CHANGED
|
@@ -34,4 +34,6 @@ The training pairs in this dataset are generated by taking each example from the
|
|
| 34 |
|
| 35 |
The articles featured in each split are disjoint from the other splits, and each split has the same number of positive (semantically the same) and negative (semantically different) examples.
|
| 36 |
|
|
|
|
|
|
|
| 37 |
Generation scripts can be found [in the GitHub repo](https://github.com/Exr0nProjects/wiki-entity-similarity).
|
|
|
|
| 34 |
|
| 35 |
The articles featured in each split are disjoint from the other splits, and each split has the same number of positive (semantically the same) and negative (semantically different) examples.
|
| 36 |
|
| 37 |
+
For more details on the dataset motivation, see [the paper](https://arxiv.org/abs/2202.13581). If you use this dataset in your work, please cite it using the ArXiv reference.
|
| 38 |
+
|
| 39 |
Generation scripts can be found [in the GitHub repo](https://github.com/Exr0nProjects/wiki-entity-similarity).
|