CHRONOS: A Schema-Based Event Understanding and Prediction System

Authors

  • Maria Chang IBM Research
  • Achille Fokoue IBM Research
  • Rosario Uceda-Sosa IBM Research
  • Parul Awasthy IBM Research
  • Ken Barker IBM Research
  • Sadhana Kumaravel IBM Research
  • Oktie Hassanzadeh IBM Research
  • Elton Soares IBM Research
  • Tian Gao IBM Research
  • Debarun Bhattacharjya IBM Research
  • Radu Florian IBM Research
  • Salim Roukos IBM Research

DOI:

https://doi.org/10.1609/aaai.v38i21.30323

Keywords:

Natural Language , Knowledge-Based Systems , Track: Emerging Applications

Abstract

Chronological and Hierarchical Reasoning Over Naturally Occurring Schemas (CHRONOS) is a system that combines language model-based natural language processing with symbolic knowledge representations to analyze and make predictions about newsworthy events. CHRONOS consists of an event-centric information extraction pipeline and a complex event schema instantiation and prediction system. Resulting predictions are detailed with arguments, event types from Wikidata, schema-based justifications, and source document provenance. We evaluate our system by its ability to capture the structure of unseen events described in news articles and make plausible predictions as judged by human annotators.

Downloads

Published

2024-03-24

How to Cite

Chang, M., Fokoue, A., Uceda-Sosa, R., Awasthy, P., Barker, K., Kumaravel, S., Hassanzadeh, O., Soares, E., Gao, T., Bhattacharjya, D., Florian, R., & Roukos, S. (2024). CHRONOS: A Schema-Based Event Understanding and Prediction System. Proceedings of the AAAI Conference on Artificial Intelligence, 38(21), 22871-22877. https://doi.org/10.1609/aaai.v38i21.30323