As Machine Learning becomes more prominent in the military, we are faced with a different take on the old problem of how to collect data relevant to some military mission need. We can now embrace the paradigm of too much data where previously we needed to focus on data reduction because humans can only process a finite amount of information. Commanders, analysts, and intelligence officers are often tasked with understanding the current situation in a mission area to create a common operating picture in order to complete their mission objectives. Data pertaining to missions can often be scraped from multiple domains, including patrol reports, newswire, and RF sensors, image sensors, and various other sensor types in the field. In this paper, we describe a system called the Multi-Domain Integration and Correlation Engine (MD-ICE), which ingests data which ingests data from two domains: textual open source information (newswire and social media)and sensor network information, and processes it using tools from various machine learning research areas. MD-ICE manipulates the resulting data into a machine readable unified format to allow for labelling and inference of inter-domain correlations. The goal of MD-ICE is to utilize these information domains to better understand situational context, where open source information provides semantic context (i.e. what type of event, who is involved, etc...)and the sensor network information provides the fine-grain detail (how many people involved, exact area of the event, etc...). This understanding of situational context in turn can, with further research, help commanders reach their mission objectives faster through better situational understanding and prediction of future needs.