Towards A Smarter Online Debates Analysis: A French Public Portal Use-Case

Ghislain Atemezing (1) , Hacene Cherfi (1) , Anh Huynh (1) , Martin Coste (1)

1 MONDECA R&D - Mondeca Recherche & Développement

This paper describes a semantic annotation workflow and visualization of politician speeches from the French Government Portal. The pipeline consists of an ontology modeling, a pre-processing and cus-tomized annotations of 7200 random speeches of public actors during the last 30 years using Content Augmentation Manager (CA-Manager) and a visualization component based on the central notion of a Topic. The implementation of the proof-of-concept shows that adding semantic annotations to the current publication of official speeches in RDF leads to efficient search of debates and better interconnection with external datasets.

Mots clés: speeches; debates; semantic annotations; interoperability; TopicPages; ITM; CAM-Manager

Domaines: Intelligence artificielle [cs.AI]; Web; Ingénierie assistée par ordinateur

Ghislain Atemezing, Hacene Cherfi, Anh Huynh, Martin Coste. Towards A Smarter Online Debates Analysis: A French Public Portal Use-Case. [Research Report] Mondeca. 2015. ⟨hal-01195604⟩