The World Health Organization first reported an irregular spike in Ebola hemorrhagic fever in Guinea on March 25, 2014. Nine days before WHO officials raised the flag, however, a website forecasted the pending outbreak, now the deadliest Ebola epidemic in history with more than 2,800 people killed so far.

Health agencies such as the WHO and the U.S. Centers for Disease Control and Prevention frequently publish health information on infectious disease outbreaks, but it is often compiled days or even weeks after an outbreak has occurred. Tracking the spread of a deadly disease as it is happening is a tricky process, and health specialists around the world are anxious to better predict when and where an outbreak could hit. While advances in satellite imaging and better weather models have made plotting an outbreak easier, experts say that the best tool available to health officials for predicting an outbreak is the Internet.

Aside from the Internet’s obvious benefits of faster communication and lightning-speed access to information, artificial intelligence software can now process data from tens of thousands of Web pages -- and in multiple languages -- hourly, according to Scientific American. A host of newly launched disease-monitoring websites such as HealthMap, which is the one that beat the WHO to predict the current Ebola outbreak, combine information from informal sources such as social media and local news outlets with official government and health websites, as well as other sources such as e-mails and on-scene accounts, to track disease outbreaks, the Associated Press reported.

These websites improve real-time outbreak coverage by using algorithms to devour, digest and make sense of thousands of online sources simultaneously and faster than any health agency or international watchdog could. There are even programs that identify the most heavily trafficked airports and connections across the globe to help predict the most likely places a particular disease is to travel. HealthMap’s algorithms can sort data by country and even account for different control scenarios.

When filtered correctly, social media can be an extremely useful tool for pinning down disease cases in an outbreak, experts say. Twitter provides some context and description of what is taking place on the ground. Additionally, tweets record the location and time of posts, and Twitter profiles contain demographic information like age and gender. A study from 2011 found that social media can be a fairly accurate gauge of health issues. Another study from 2010 found that social networks can predict flu outbreaks much earlier than traditional tracking methods.

By the time the WHO formally recognized the current Ebola outbreak, 86 cases had been reported, including 59 deaths. The outbreak has since spiraled out of control and is well on its way to infecting some 1.4 million people across West Africa by January.

Although the current outbreak of Ebola is not the first major health crisis that HealthMap has covered, advances in HealthMap’s software have improved the early-warning system. Also, other health-monitoring websites such as Google Flu Trends and the Global Public Heath Intelligence Network could provide additional context in the face of an outbreak. “It shows some of these informal sources are helping paint a picture of what’s happening that’s useful to these public health agencies,” HealthMap co-founder John Brownstein told the AP in August.