Twitter

Previous research identifies social media as an informal source of near-real time health data that may add value to disease surveillance systems by providing broader access to health data across hard-toreach populations. This indirect health monitoring may improve public health professionals’ ability to detect disease outbreaks faster than traditional methods and to enhance outbreak response. The Philippines consists of over 7,000 islands and is prone to meteorological (storms), hydrological (floods), and geophysical disasters (earthquakes and volcanoes). In these situations, evacuation centers are used for safety and medical attention and often house up to 50K people each for 2 or more months, sometimes with unclean water sources and improper sanitation. Consequently, these conditions are a perfect venue for communicable disease transmission and have been proposed to cause disease outbreaks weeks after the original disaster occurred. Coined the social media capital of the world1, the Philippines provides a perfect opportunity to evaluate the potential of social media use in disease surveillance.

Objective

To determine the potential of Twitter data as an early warning of a likely communicable disease outbreak following a natural disaster, and if successful, develop an open-source algorithm for use by interested parties.

 Numerous methods using social media for syndromic surveillance and disease tracking have been developed. Many websites use Twitter and other social media to track specific diseases or syndromes.1 Many are intended for public use and the extent of use by public health agencies is limited.2 Our work builds on 4 years of experience by our multi-disciplinary team3 with a focus on local surveillance of influenza. 4,5

Objective

Create a flexible user-friendly geo-based social media analytic tool for local public health professionals. With the goal of increasing situational awareness, system has capability to process, sort and display tweets with text terms of potential public health interest. We continue to refine the Social Media and Research Testbed (SMART) via feedback from surveillance professionals.

Traditional influenza surveillance relies on reports of influenzalike illness (ILI) by healthcare providers, capturing individuals who seek medical care and missing those who may search, post, and tweet about their illnesses instead. Existing research has shown some promise of using data from Google, Twitter, and Wikipedia for influenza surveillance, but with conflicting findings, studies have only evaluated these web-based sources individually or dually without comparing all three of them1-5. A comparative analysis of all three web-based sources is needed to know which of the web-based sources performs best in order to be considered to complement traditional methods.

Objective

To comparatively analyze Google, Twitter, and Wikipedia by evaluating how well change points detected in each web-based source correspond to change points detected in CDC ILI data.

Social media messages are often short, informal, and ungrammatical. They frequently involve text, images, audio, or video, which makes the identification of useful information difficult. This complexity reduces the efficacy of standard information extraction techniques1. However, recent advances in NLP, especially methods tailored to social media2, have shown promise in improving real-time PH surveillance and emergency response3. Surveillance data derived from semantic analysis combined with traditional surveillance processes has potential to improve event detection and characterization. The CDC Office of Public Health Preparedness and Response (OPHPR), Division of Emergency Operations (DEO) and the Georgia Tech Research Institute have collaborated on the advancement of PH SA through development of new approaches in using semantic analysis for social media.

Objective

The objective of this analysis is to leverage recent advances in natural language processing (NLP) to develop new methods and system capabilities for processing social media (Twitter messages) for situational awareness (SA), syndromic surveillance (SS), and event-based surveillance (EBS). Specifically, we evaluated the use of human-in-the-loop semantic analysis to assist public health (PH) SA stakeholders in SS and EBS using massive amounts of publicly available social media data.