Graphically construct and execute rich data analysis queries over Wikidata and other RDF Data Sets.
 |
The paper: "Towards UML-style Visual Queries over Wikidata"
Publication by Springer: Towards UML-style Visual Queries over Wikidata
Local authors' manuscript:
Google Chrome browser is recommended (may work also in other browsers).
Register (sign up) as a new user (email checking not done, any e-mail-like string is accepted), then log in.
Create a new project (give at least the name for the project), initialize your project by the predefined wikidata project configuration ('Initialise by wikidata sample project'). If an empty project is created, its contents can be uploaded (button 'Upload Project') later from a sample project file or URL, as e.g., Wikidata sample project.
Create a new diagram, click on the diagram pane. The tabs for classes, properties and individuals shall be to the right. Double-click an item in a list to seed a query from it.
The project file can be separately downloaded (in JSON-based abstract syntax).
Analysis of wikidata example SPARQL query visualizations (.pdf)
The success rate of automated visualization along wikidata example SPARQL query categories:
| Category | Success | Total | Percent |
| Simple Queries | 21 | 21 | 100 |
| Lexeme Queries | 17 | 17 | 100 |
| Wikibase Predicates | 7 | 8 | 87.5 |
| Wikimedia projects | 13 | 13 | 100 |
| Entertainment | 17 | 20 | 85 |
| Computer Science and Technology | 11 | 11 | 100 |
| Biology and Medicine | 25 | 30 | 73.33 |
| Total | 111 | 120 | 92.5 |
The code for data shape server (serving the names for auto-completion) is in its own repository on GitHub.
The code and instructions for retrieving and storing wikidata schema is in the wikidata folder within the data shape retrieval services repository.