Purdue niversity researchers have developed an algorithm that analyzes tweets to better understand how users respond to crises, offering a new way to inform decisions on disaster management.
The algorithm captures individuals' responses to a particular crisis via their tweets, creating a "fingerprint" that researchers and policy makers can use to better quantify what the community thinks about that crisis.
The technique examines individuals' tweets associated with an event and breaks them down into categories of community resilience: ecological, economical, institutional, social, infrastructure, and quality of life.
The algorithm automatically calculates and generates a heat map of the fingerprint, which makes it easier to determine which aspects of community resilience are more prevalent in people's reactions to certain crises.
From Purdue niversity News
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, SA