A unique component of DT’s robust text mining platform is our Event Extraction Engine. We have the ability to recognize and extract events from within unstructured text. We can extract these events from news stories with both high precision and high recall.
Using our CaRE platform, we can quickly develop additional event types relevant for any customer-specified domain, with higher accuracy than other alternatives. To date, users have primarily focused on financial events and we have defined and coded the following types:
The event extraction engine has various applications, and is currently used in conjunction with our Sentiment Analysis Engine to power The Stock Sonar financial sentiment analysis application. By identifying events in text, the sentiment can be properly understood in context. In addition, we have created a Significant Developments news application where we can apply customer-specific scoring rules to define what makes a news story and event significant.
For more information on our Event Extraction Engine, please contact us at email@example.com.