Using Text Analytics To Detect Animal Diseases

The Challenge

A federal agency tasked with protecting America’s agricultural infrastructure and economy from emerging biological threats wanted to track and respond to infectious animal disease events to ensure national security and economic stability. The goal was to enhance agency analyses using real-time, state-of-the-art tools that could acquire, assess, and integrate comprehensive data from the internet and commercial, government, and Non-Governmental Organizations databases.

The Solution

Elder Research combined state-of-the-art data mining and machine learning tools with best practices for data integration to enhance analysts’ efficiency in developing country-specific veterinary capability studies. A visual representation of the system is shown below. The proprietary solution employed cutting-edge “learned rule” natural language processing, which did not rely on dictionaries and heuristics.

Available technologies and software tools were evaluated to determine the most cost-effective solution for the real-time data streaming architecture and the best way to provide access to the results. Based on input from the client the system was designed to take advantage of analysts’ expertise in many different forms and employ multiple methods for gathering information. Those documents with the highest ratings are presented to the analyst as the most likely documents of interest. Intuitive tagging and scoring features enabled analysts to supply feedback on documents, enabling the system to truly learn and improve system performance.


Automating the analysis of text data from reliable sources made the work of sifting through huge databases and incoming information streams manageable, intuitive, and efficient. Reliable event reports enabled analysts to validate incidents of animal disease and make recommendations for dealing with them at the earliest possible stage.

Download This Case Study