Related Experiment Videos
Extending the event calculus for tracking epidemic spread
1Laboratoire d'Informatique Fondamentale, UMR CNRS 6166, Université de la Méditerranée, Faculté de Médecine, 27 bd Jean Moulin, 13385 Marseille Cedex 5, France. herve.chaudet@medecine.univ-mrs.fr
Artificial Intelligence in Medicine
|August 4, 2005
Summary
A new formal language, SpatioTemporal Extended Event Language (STEEL), was developed to represent disease outbreak histories from reports. This system demonstrated agreement with human experts in tracking epidemic spread.
Area of Science:
- Epidemiology
- Computational Linguistics
- Artificial Intelligence
Background:
- Email reporting systems are crucial for early disease outbreak detection and surveillance.
- Tracking epidemic spread requires effective representation of outbreak histories.
- Existing methods may lack the precision needed for detailed spatio-temporal analysis of disease events.
Purpose of the Study:
- To develop a formal language for event-centered representation of disease outbreak histories.
- To enable tracking the spread of epidemics using structured outbreak reports.
- To create a system for analyzing the spatial and temporal dynamics of disease outbreaks.
Main Methods:
- Developed the SpatioTemporal Extended Event Language (STEEL), an extension of Event Calculus.
- STEEL focuses on joint spatial and temporal event occurrences and event aggregation.
- Implemented STEEL in Prolog and hand-coded a corpus of 35 outbreak reports.
Main Results:
- The STEEL system was tested on a corpus of outbreak reports.
- Performance was compared against three human experts in a question-and-answer task.
- The system showed agreement with human expert responses, validating its capabilities.
Conclusions:
- STEEL effectively captures the spatial and temporal location of event occurrences in outbreak reports.
- The language's representation closely mirrors narrative descriptions.
- Future work will focus on automating the modeling of outbreak reports using STEEL.