Predicting the future with wikidata and wikipedia

Abstract

In this demonstration, we present a prototype knowledgebased event forecasting system powered by Wikidata and Wikipedia. The system relies on existing event-related concepts and relations in Wikidata to build a base knowledge graph of events and consequences. It then uses a combination of state-of-The-Art knowledge extraction methods to augment the base knowledge graph using natural language descriptions of events and their consequences that can be found in Wikipedia articles. Using a number of use case scenarios, we show how the resulting knowledge graph can be used as a part of a human-in-The-loop explainable solution for event forecasting and analysis.