Datashare and Neo4j: From Unstructured Documents to Knowledge Graphs for Investigative Journalism

Welcome to NODES 2024

Welcome to NODES 2024, the biggest graph community gathering dedicated to graph-powered apps and contextual AI with the latest GenAI innovations!

Join us on November 7, 2024, for a 24-hour free online conference accessible across all time zones and locations, catering to all levels of expertise.

Session Track:

Session Time: 08:00 am - 08:15 am November 7

Session description

Have you ever wondered how journalists extracted valuable information from the millions of documents of the Panama Papers or Pandora Papers? In this session, Clément, ML Engineer at the International Consortium of Investigative Journalists (ICIJ), will explain how Neo4j and Datashare (ICIJ’s search engine) helped solve this challenge. You will learn how Datashare’s new Neo4j plugin allows journalists to turn investigation documents into knowledge graphs, connecting the dots between people and corporate entities. You will then learn how to use Neo4j to uncover valuable insights from such graphs to reveal complex links between entities, identify central actors, analyze exchanges, and identify similar entities.

Speaker

Clément Doumouro

Machine Learning Engineer, ICIJ

Clément Doumouro is an ML Engineer at the International Consortium of Investigative Journalists (ICIJ). After a Neo4j Fellowship focusing on connecting the dots between entities found in investigation documents, he joined ICIJ permanently to tackle Natural Language Understanding and ML engineering challenges. Clément holds a master's in AI and robotics from Centrale Supélec. He previously spent seven years at Sonos, building the core of the natural language understanding technology powering the Sonos Voice Control, a privacy-preserving, on-device speech recognition technology embedded on Sonos speakers.