Interactive exploration of complex relational data sets in a web



With the increase of large inter-linked open data sets made available on the web, there is a growing interest in tools that allow to quickly and easily store, transform, query, mine and visualize that data.

In this talk, we focus on our use of the Protovis,D3 Javascript library to interactively visualize the content of a relational database. The underlying framework, CubicWeb <>_, is written in Python and relies on the numpy and scipy libraries for the intensive numerical computations.

CubicWeb is a semantic web framework written in Python that has been succesfully used in large-scale projects, such as Data.bnf (French National Library’s opendata) or Collections des musées de Haute-Normandie (museums of Haute-Normandie). Using a browser connected to the server via HTTP, the user can enter queries in a high-level query language, similar to SPARQL but called RQL, that operates over a relational database PostgreSQL in our case.

Data will be loaded from Geonames, DBpedia, various RSS feeds and

Using Protovis, views will include maps, charts, hierarchies, networks, statistics, etc. A important feature is that any tuple (query, processor, view) has a corresponding url, making all results addressable, linkable and shareable.

More technical details can be found in this blog post : "Data Fast-food": quick interactive exploratory processing and visualization of complex datasets with CubicWeb.

Fichiers joints