This is the FB15k-CVT dataset, a knowldge graph that exapnds the original FB15k-237 dataset. ...

The dataset is organized into two main folders :

• Knowledge Graphs: contains the original FB15k-CVT knowledge graph (FB15k-CVT.nt) in N-Triples format. Also contains FB15k-CVT_Aux.nt KG which is an extension of FB15k-CVT with additional entities (defined with MIDs) in relation with entities of type CVT.
• Stratified Dataset: contains the stratified train/validation/test splits of the FB15k-CVT dataset, as well as the quadruples used for the evaluation.

The folder structure is as follows:

+ KnowledgeGraph
|--- FB15k-CVT.nt
|--- FB15k-CVT_Aux.nt 
+ Stratified
|--- train.nt
|--- valid.nt
|--- test.nt
|--- +Quadruples
       |--- valid_t1.csv
       |--- valid_t2.csv
       |--- valid_t3.csv
       |--- test_t1.csv
       |--- test_t2.csv
       |--- test_t3.csv

• train.nt, valid.nt, and test.nt contain the triples for the train, validation, and test sets, respectively, in N-Triples format.
• Quadruples contains the quadruples used for evaluation, split into three sets (t1, t2, and t3, that) for each of the validation and test sets. Each quadruple is represented as a CSV file with four columns: (Entity_01, relation_01, relation_02, Entity_02). Each set corresponds to a different type of evaluation scenario:
  • Direct paths (t1): These quadruples are used for evaluating the model on (i) chain backward prediction task i.e., (?, 𝑟1, 𝑛_cvt)/(𝑛_cvt, 𝑟2, 𝑒2), (ii) chain forward prediction task i.e., (𝑒1, 𝑟1, 𝑛_cvt).
  • Splitting paths (t2): for (iii) join prediction task i.e., (𝑒1, 𝑟1, 𝑛cvt)/(?, 𝑟2, 𝑛cvt)
  • Joining paths (t3): for (iv) split prediction task i.e., (𝑛cvt, 𝑟1, 𝑒1)/(𝑛cvt, 𝑟2, ?)

If you use this dataset in your research, please cite the following paper:

Mouloud Iferroudjene, Victor Charpenay and Antoine Zimmermann. (2023, July). "FB15k-CVT: A Challenging Dataset for Knowledge Graph Embedding Models." In NeSy 2023, 17th International Workshop on Neural-Symbolic Learning and Reasoning.

Or,

@inproceedings{iferroudjene2023fb15k,
  title={FB15k-CVT: A Challenging Dataset for Knowledge Graph Embedding Models},
  author={Iferroudjene, Mouloud and Charpenay, Victor and Zimmermann, Antoine},
  booktitle={NeSy 2023, 17th International Workshop on Neural-Symbolic Learning and Reasoning},
  year={2023}
}

This dataset is released under the Creative Commons Attribution ...