Arvind Chandrasekaran, Colorado Technical University, USA

Hints of the outbreak are detected through the modified circumstances favoring the outbreaks, like the warm weather contributing to epidermal outbreaks or the loss of sanitation leading to cholera outbreaks typically relying on the routine reports from the healthcare facilities, secondary data like attendance monitoring at workplaces and schools, the web, and the media play a significant informational source with more than 60% of the initial outbreak reporting to the informal sources. Through the application of natural language processing methods and machine learning technologies, a pipeline is developed which extracts the critical entities like country, confirmed case counts, disease, and case dates, which are mandatory entities from the epidemiological article and are saved in the database thereby facilitating the data entry easier. The advantages are the facilitation of relevant score articles shown first, thereby providing the web service results termed EventEpi integrated into the Event Based Surveillance (EBS) workflows.

Event-based surveillance; Contribution; Methods; Information extraction; Entity filtering; Scoring; Evaluation