Skip to content

Latest commit

 

History

History
63 lines (47 loc) · 2.94 KB

WEBQSP_WD_README.md

File metadata and controls

63 lines (47 loc) · 2.94 KB

WebQSP-WD

v1.0, 2018-08

This data set contains a version of the WebQSP data set mapped to Wikidata. Compared to the original data set, this resource only includes question for which at least some acceptable answer exists in Wikidata. The test partition also includes the mapping of the output of an SMART-S entity linker that was originally produced for the WebQSP data set (see Yih et al. 2016 for details).

Please consult the corresponding code repository and the paper to learn more about how the data set was constructed and used.

WebQSP-WD is available at following location: https://public.ukp.informatik.tu-darmstadt.de/coling2018-graph-neural-networks-question-answering/WebQSP_WD_v1.zip

The input folder contains the train and test partitions with answer ids mapped to Wikidata. The generated folder contains automatically generated candidate graphs for the train partition that are needed to train a new model. Please consult the paper for mode details.

Please use the following citation:

@InProceedings{C18-1280,
  author = 	"Sorokin, Daniil
		and Gurevych, Iryna",
  title = 	"Modeling Semantics with Gated Graph Neural Networks for Knowledge Base Question Answering",
  booktitle = 	"Proceedings of the 27th International Conference on Computational Linguistics",
  year = 	"2018",
  publisher = 	"Association for Computational Linguistics",
  pages = 	"3306--3317",
  location = 	"Santa Fe, New Mexico, USA",
  url = 	"http://aclweb.org/anthology/C18-1280"
}

Reading in the data:

All data set files are in json format. For the files in the generated folder, please use the following code snippet. Make sure to download the project first.

import json
from questionanswering.construction.sentence import sentence_object_hook

training_dataset = []
with open(path_to_train) as f:
    training_dataset = json.load(f,  object_hook=sentence_object_hook)

print("Graphs for the first question: ", training_dataset[0].graphs)

You do not need a Wikidata endpoint and the additional internal projects to read in the files.

Contacts:

If you have any questions regarding the code, please, don't hesitate to contact the authors or report an issue.

License

  • This data set is derived from the original WebQuestions data set and its subset WebQSP. WebQuestions was released under CC-BY 4.0.
  • Please cite our paper if you use the data in your work.