In times of crisis, be it a war, some natural disaster or deadly sickness, one often wonders whether or not such events could have been mitigated, or better still, avoided altogether with the help of our latest technological advancements. Could it be that the answer lies with big data?

Perhaps, just perhaps, there’s a chance that it could. Take military conflict as an example. Wars rarely, if ever, kick off without any kind of warning signs or indicators that some kind of crisis is brewing. Generally, wars are the culmination of months, if not years of tension, marked by increasing rhetoric from the instigators as the situation spirals towards violence.

In the future, it might well be possible to spot these warning signs at an early stage, giving diplomats a chance to step in before tensions reach boiling point and war becomes inevitable. Working with the United Nations Development Program (UNDP), the web data analysis firm Recorded Future has developed a tool that can analyze the political environment in any part of the world via news reports, social media and other sources, in order to detect possible ‘anomalous patterns’ that could indicate an escalation of tensions before it actually happens.

The theory is that the more ‘negative’ press a country receives (i.e. mentions of military or security issues), the more likely it is that conflict could arise. Positive press meanwhile, for example statements on such things as trade, investment and tourism, would indicate that imminent conflict is most unlikely.

To test the robustness of the software, Recorded Future and the UNDP recently applied it to the four month period leading up to the border conflict between Russia and Georgia in 1998. The results of this project were discussed at length in a recent blog post, but as predicted, the software was able to spot an anomaly as Georgia was mentioned with growing frequency in the media as the outbreak of war approached.

Recorded Future later refined their search, based on the hypothesis that “statements about violence would be the key leading indicators of future violence,” as opposed to statements about such things as ‘trade and investment’ or ‘tourism’, which would indicate that violence was unlikely to occur.

Specifically, the researchers looked for any messages in the news that “contains references to military and security issues (as opposed to tourism development), focuses on response and reaction to what others are saying, it is indicative of a speaker’s discontent with the status quo”.

As they hoped, the data proved the researchers right, with statements of violence sky rocketing in the months preceding the war, whilst positive statements all but disappeared from the landscape.