The cocktail dataset can be downloaded here:
To access the creation page, check Introduction and Access section!
Pick the name and language of your QA System, upload your dataset (click or drag and drop on the dedicated area), you can click on:
- Settings to personnalize them (default parameters are good for general use)
- Finish to create your QA System (and use default parameters)
You can also modify the list of properties that are used to assign names to the nodes.
The dataset can be in one of the common RDF serializations and must be syntactically correct. You can both check and clean a file using serdi
serdi -i ntriples -o ntriples dump.nt > dump_clean.nt
After uploading the dataset, your QA system will appear in your home page.
It is now ready to be queried!