Technology Was the Only One to Hit The USA Election Prediction
Mr Stratocaster, PostDoc.
"On the 6th of January 2017 the victory of Donald Trump by the American Congress was officially declared. It was the most balanced election of all time, and the most difficult to predict.
There is a real prediction industry before the election. But in 2016 a curious fact happened. All research institutes failed in their predictions. Only two research institutes in the US pointed to the possibility of a Republican victory (IBD/TIPP Tracking and LA Times/USC Tracking). But even these made a mistake, because they projected this victory in the "popular vote," which is the direct vote of the voters. But in this scenario the Democrats won the election. The Republican Party won only in the Electoral College (state delegates). So even those who "hit", at the end also "was wrong".
The most correct projection came from a technology that uses semantic technologies and ontology engineering. A digital traffic reading report posted online at 08:02 a.m. on the 8th of November 2016, hit the 2 scenarios: Democrats would win direct voting, and Republicans would win in most states. This was the assertion made by the digital intelligence company Synaption, based on monitoring and interpretation technologies for large data flow. And it seems to have been the only source of information that correctly pointed out the main trends:
-Democrats led in direct voting;
-Republicans led in most states;
-Republicans would win in Florida with close advantage;
-The States of Michigan, Pennsylvania, and Wisconsin were tied;
-There would be a breakdown of delegates in the state of Maine, with 3x1 for Democrats;
-At that moment, 39 States were out of the margin of error, 6 states were tied, and 6 were in technical tie state.
All these data proved to be correct in determining the election (including the percentage in the state of Maine) even considering the 7 delegates who did not vote in the electoral college, and that will not be considered for statistical purposes, prevailing the result verified at the polls.
The Digital Intelligence Report - D.I.R. - released by the company showed that the projection - of 29 States to Republicans - would amount to 56.86% of the Delegates. The final result was 306 delegates conquered by the Republicans which correspond to 56.87%. A difference of only 0.01%. The D.I.R. also predicted that the final difference between Democrats and Republicans would be 2% in popular vote. The final result was 2,1%, a difference of only 0,1%.
But how is this possible? The answer lies of the use of innovative inference and behavioral interpretation technologies based on Ontology Management (holonymy, meronymy, hyperonymy, hyponymy, synonyms and related terms monitoring). This allows one to determine the course of events more accurately than traditional surveys.
The origin of the technologies lies of the experience of the company's creation team, which previously developed systems in the area of intelligence and investigations, wiretapping, border control systems, knowledge bases and similar. The company has mastered this kind of technology to search engines, and has been using it to map behavior of large masses of data on the Internet.
Based on these technologies, the company has just provided two new thematic search engines, one in the area of Law (Synaption Juris http://www.synaption.com/juris/), and another in the area of Medicine and Health (Synaption Med http://www.synaption.com/med/). The new search engines employ a technology that does a context reading, and when the user changes the context of the question the search result changes automatically.
Currently, the company has just launched a portal called "Mr. Predictions" (http://www.mrpredictions.com) which is based on these technologies. It features scenarios readings of the Stock Exchange and Futures Markets, oil market monitoring, sports and election predictions. Using the site technology it is possible to know, for example, if the chance of a world war is growing or decreasing and what the impact of this information on the markets and society is."
(*) Statistical data about the US election is available at:
(2) Predictions conducted for the American Election:
(3) Digital Intelligence Report about US 2016: