Thesis of Sarra Ouelhadj


Subject:
Definition of a sustainable workflow for semantically enriching territorial Open Data

Summary:

The Métropole de Lyon publishes Open Data datasets produced by itself or its partners via the data.grandlyon.com platform. Most of the published data are available in standard formats (e.g. CSV, JSON, GeoJSON, etc.) ensuring a good level of syntactic interoperability. However, the semantic aspects are treated in a rather superficial way. The meaning of the data is often transmitted "out of band", via textual documentation in the metadata, or via private communications between data consumers and the platform's user support. In addition, consumers sometimes rely on their own interpretations, which may be wrong. As a result, much of the value of published Open Data is difficult to exploit.

The notion of "Linked Data" denotes a set of technologies and standards recommended by the W3C, aimed at resolving these problems mentioned. It is based on a directed edge labelled graph data model (RDF) that facilitates data integration and the use of shared vocabularies with explicit semantics. 

The Métropole de Lyon intends to make its Open Data progressively conform to the principles of Linked Data, the thesis work will focus on defining a migration process that is robust to changes in the original data, and to enable the teams at the Métropole de Lyon to identify the changes to be made to the process and the best way to implement them.


Advisor: Stéphanie Jean-Daubias
Coadvisor: Pierre-Antoine Champin