• Marcel Ramirez


I am going to try to explain, in a very simple and didactic way, how data mining has changed the way of planning a political campaign or defining government strategies.

You must first understand what data mining is and how its complex interaction with machine learning changed the rules of the game. Data Mining is essentially the collection of data on a massive scale, while machine learning is a subset of artificial intelligence that relies on teaching computers to "think" like human beings so that they can solve abstract problems.

GOOGLE is a great example of data mining and machine learning in action. Their algorithms can process the huge amounts of search data they store from each of their users and then process the numbers and make super relevant recommendations. The machine learning algorithm learns as it goes, which means the more data you have access to, the better. That is why you will be able to realize that the search recommendations on GOOGLE or the information you receive from the search engine are always related to your search topics.

Initially, it may seem to us that this is not related to politics, but the same principle applies no matter what the information is about. So, for example, imagine if the office of the Director of Transit or Secretary of Mobility had access to real-time traffic data that could be analyzed using machine learning algorithms to provide, through artificial intelligence, real-time suggestions on when to close the roads. or redirect traffic; we are talking about an algorithm that has the potential to solve the mobility problem faced by large cities in the world.

The power of data

Data is knowledge and knowledge is power, one of the reasons why data mining has changed the way we think about politics.

Interpreting the data can make or break a political campaign. The reality is that they can help people get elected to a position and do their jobs much more effectively and efficiently.

We have already talked about the data that is used to improve traffic flows and make roads safer. Now imagine that the same concept could be implemented in every area that is the government's job to oversee and facilitate.

For example, healthcare managers can use the data and its analysis to determine where to best allocate funds. It can be used by foreign ministers to simulate complex trade deals or to predict the long-term effects of uncertain political situations, such as the UK's decision to leave the European Union. It can be used to identify potential terrorist threats or to give advance warnings of disease outbreaks or other phenomena using population data.

Data mining in political campaigns

Charisma, leadership, but also Data Mining and the analytical skills to exploit it. These seem to be the perfect match for the winning candidate based on what emerges from the latest electoral experiences around the world. Barack Obama's reelection campaign was the beginning of what would be the new era of digital strategies in the world.

The re-election of Barack Obama, the first 100% digital campaign.

Obama's reelection campaign, Trump's 2016 and 2020 campaign, Biden's 2020 campaign, etc. they went beyond the usual use of voter databases, classified according to the classical subdivision into sociodemographic categories. Thanks to the analysis of the information obtained from the data collection, the campaigns were able to evaluate on the basis of the preferences and habits expressed online by each voter. The application of data mining techniques made it possible to become aware of people's propensity to vote, towards which a meticulous microtargeting campaign was directed.

But definitely the King of data mining campaigns as the undisputed protagonists in the construction of a communication strategy for political use, is that of the 2020 presidential elections in the United States, where we could see a strategic war between Trump and Biden based exclusively on data analysis.

While Biden's campaign focused on identifying undecided voters who could go either way in areas with a high probability of voter turnout, Trump's campaign focused on choosing his key political platforms; for example, reducing illegal immigration would work best with segmented constituencies. Once the 2020 reelection analytics system was up and running in late summer 2020, the campaign sent personalized messages to 1,000,000 targeted voters every day.

Using sophisticated analytical techniques and a surgical approach to digital marketing, a candidate can convey his electoral messages in the ways, times and in the most appropriate places to reach each constituent. Taking, for example, the Trump 2020 re-election campaign, on the issue of gun ownership, where it was proposed as a “defense” for scared citizens living in dangerous neighborhoods, “constitutional” for the purists of the ten amendments and as a "tradition" to lovers of sport hunting, the same message sent in different ways according to the type of voters.

These techniques have also been used by the UK Ministry of Defense, the US State Department and NATO, to understand how to neutralize Islamic recruitment online.

"We find your voters and put them into action," was the slogan of Cambridge Analytica.

Data mining inserted in a political context becomes a weapon of consensus. The method used combines psychometric analysis (the field of psychology that specializes in measuring the psychological characteristics of the individual) with data mining techniques. From this intersection, OCEAN (acronym in English) is generated, the model based on the analysis of the big five: openness, awareness, extraversion, collaborative approach and emotional stability. The analysis of these five basic variables allows to know the voters, understand their needs, fears, possible behaviors and then translate them into personalized publications for selected audiences. This also allows you in real time to be able to make a projection of voting intentions without the need to carry out a survey, imagine being just one click away from knowing how the voting preference trends are going without having to wait two or three weeks for the results of a survey.

The data necessary to achieve the success of digital technology, comes from the collection of data from the internet, this digital tool differs from traditional databases by three characteristics that must necessarily be present at the same time: large volumes, speed and variety of sources and nature. The Internet, with its 450 million daily transactions, is obviously the main source of this extraordinary resource.

Amazon, Facebook, Google, etc., use the profile generated by the data mining analysis to create effective marketing strategies, and we live with this every day, which already seems very normal to us.

A campaign based on data mining and artificial intelligence can be much cheaper than a traditional campaign and needs much less staff or collaborators. For example, for the 2016 Trump campaign, to win, less than half of the $ 521 million spent by his adversary Hillary Clinton.

The future is Data Mining in political campaigns.

0 visualizaciones0 comentarios