This project is a data visualisation class project targetting developers working with the semantic web. It is our attempt to visualise results of SPARQL queries (e.g. from dpbedia), which usually come in the form of long tables containing lots of URIs (though we actually focus on URLs, not all kinds of URIs).
The interface we propose contains two complementary parts:
- Per-column dashboards containing basics statistics (e.g the most common URL domains), keywords, and visualisations of the structures of the URLs (using a Sankey diagram). They allow to quickly read some common metrics and get the gist of each column.
- A data table containing the entire result of the SPARQL query. It is split into lazyily generated pages (for better performances and ease-of-use) and offer enhanced representations of the URLs.
The project is written in TypeScript (transpiled into plain JS for the browser) plus some HTML and CSS. It has no runtime dependency (in particular, DOM manipulation is done using vanilla JS).
We use Rollup to bundle the transpiled TypeScript files into a single JS file.
We use yarn to manage the development dependencies and to run scripts, but it should work fine using your favourite package manager for JavaScript (such as npm).
Run the following commands to create your own build:
yarnwill install the development dependencies.yarn buildwill compile and bundle TypeScript sources into thebuildfolder.
Note that the HTML and the CSS files of the demo pages are not processed by Rollup (they are located in the demo directory).
You can also run yarn watch to start Rollup in watch mode (i.e. automatically recompile the sources when they change).