Recording – FactForge Debuts: Trump World Data and Instant Ranking of Industry Leaders

FactForge Debuts: Trump World Data and Instant Ranking of Industry Leaders

This webinar has been recorded and available on demand.

This webinar continues series are demonstrating how linked open data and semantic tagging of news can be used for comprehensive media monitoring, market and business intelligence. The platform for the demonstrations is FactForge: a hub for news and data about people, organizations, and locations (POL). FactForge embodies a big knowledge graph (BKG) of more than 1 billion facts that allows various analytical queries, including tracing suspicious patterns of company control; media monitoring of people, including companies owned by them, their subsidiaries, etc. The most popular webinar of this series presented analytics on the Panama Papers dataset, which Ontotext published as LOD (at http://data.ontotext.com) and included it in FactForge.

The new data and tools, which will make their debut with this webinar, include:

  • Trump World data – a BuzzFeed project where four journalists have consolidated from public declarations of trump and investigations of other journalists. The data is available as tables of relationships between people and organizations. We RDF-ized this data with OntoRefine, organized the hundreds of different relationship types into an ontology and use GraphDB to find interesting patterns in this data. Of course, we mapped many of those people and organizations to DBPedia to be able to use additional information about them in our analysis.
  • Company ranking service, which provides popularity ranking of the companies from a given industry sector. Rankings can be customized by specifying a geographic region, news category (e.g. business, sport, lifestyle) and time period. What makes this service unique is that it not only tracks mentions of the companies in the news but also includes mentions of their subsidiaries. Unlike FactForge, this service is aimed at non-technical users as it doesn’t require knowledge of SPARQL.

From a technology perspective we will demonstrate multiple features of GraphDB and its new release 8.1: importing tabular data in RDF using OntoRefine; mapping freshly RDF-ized data to DBPedia; visual exploration of knowledge graphs. The entire FactForge platform pushes the frontiers of semantic technology in several ways: efficient reasoning with more than 1 billion facts; efficient enforcement of identifier mappings (owl:sameAs) across datasets, namely half million mappings between DBPedia and Geonames. FactForge incorporates news metadata, namely tags that link entity mentions with their identifiers and further information about them in the BKG. This news metadata is imported live from Ontotext’s NOW semantic news portal that includes about 2000 articles of general news in English, automatically tagged with the Dynamic Semantic Publishing platform.

Who is this training for:

  • People with interest in one of the following: market intelligence, open data, investigative journalism;
  • Good for both technical and non-technical audience.

Join this webinar to learn how you could utilize Ontotext’s platform FactForge.

Webinar Duration: 40 minutes + 10 minutes for questions

Live Webinar: Yes

Live Question Answering: Yes

Resources After Webinar: Recording

Your experienced tour guide throughout the session is Atanas Kiryakov. Atanas is the founder and executive director of Ontotext AD. He joined Sirma as a software engineer in 1993 and later on became partner and member of the board. In 2000 he founded Ontotext. Since 2004 Atanas is a member of the board of Innovantage – a leading provider of recruitment intelligence in the UK.

Back to top