Strategic Foresight Automation Prototype “Horizon” (2019)

Goal:
To create a tool that provides a systematic way of looking for activities shared on social media platforms that may be relevant to public health. The tool can also analyze emerging trends and perform in-depth exploration of specific terms of interest.

Summary:
The goal was to create a web application that could analyze social media text to detect and explore potential emerging trends that may be relevant to public health. The system continually ingests Reddit information, scans for public health-related content, and then extracts the most relevant or popular terms being discussed. In addition, the tool provides an in-depth way for researchers to explore Reddit data for specific search terms of interest. This project demonstrates the potential use of natural language processing (NLP) and machine learning (ML) technologies to help reduce blind spots to public health hazards and create an innovative early detection methodology to spot emerging trends. The result is a tool that can easily be referenced by a public health policy team to explore trends occurring in public conversation. Horizon helps public health experts convert a massive repository of unformatted text data into a wealth of relevant information. This simple tool helps maximize the efficiency of trend detection efforts, allowing researchers to quickly see through the noise and notice important patterns that could affect the broader population.