How Machine Learning Allows the FBI to Analyze Nearly 8000 Emails Per Day

How Machine Learning Allows the FBI to Analyze Nearly 8000 Emails Per Day
This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.
<p>Machine Learning</p>

Machine Learning

Machine learning also called “deep learning ” is an advanced computational method that has changed the way we use technology. Both Google and Facebook have made this technology available to consumers with their search and social technologies. While the FBI plans to keep their software private, they’ve demonstrated just how powerful this technology can be.

What is Machine Learning?

When we write instructions for our computers, those instructions are followed exactly as we wrote them. If the programmers didn’t prepare for certain conditions to be met, the program won’t run. It will just fail with an error.

Instead of blindly following a set procedure, machine learning allows the computer to measure the outcome. To illustrate how this works, let’s look at how this applies to a simple task. When you write a program, you’re explaining a procedure to a computer. At first, glance, cleaning your kitchen seems like a pretty simple procedure. You wash the dishes, clean off the counters, and then sweep the floor. But you don’t always follow the same procedure. If you’re having trouble scrubbing a dish, you might decide that it would be more efficient to let the dish soak for a few hours, then try again. People learn from their experience. Today, you can probably tell when a dish needs to be soaked just by looking at it. You know this because, at one point, you’ve put tons of energy into scrubbing that frustrating dish by hand.

Machine learning allows computers to learn from their experiences just like we do. In the grand scheme of things, this means that they can take complex tasks and develop algorithmic ways to address them in the most effective manner possible.

Technology’s Role in Society

The software is only valuable when it can be used to improve our lives. Google’s Deep Dream, one of the first public examples of machine learning, was certainly fun to look at. But the program wasn’t much more than a novelty. They’d proved how powerful the system could be, but we need a way to apply it to improve our lives. The presidential election provided us with the perfect opportunity.

Transparency is one of the most important parts of an effective democracy. Politicians can only be effective if they are actively working with other areas of the government to make important decisions. This can only be verified if there is transparency. As part of this transparency, it’s important that anyone with upper-level roles in the government direct their digital communication through secured government servers. Not only is this a matter of national security, but it ensures that the FBI is able to investigate potential corruption. Donald Trump accused her of rigging the election, and the fact that she had been using a secret, unapproved email server raised a few eyebrows. Once this was discovered, it was the FBI’s job to ensure that she wasn’t attempting to hide any underhanded activity.

How Machine Learning Solved an Important Political Issue

There were over 650,000 emails located on this server. Even with a team of trained investigators, it would surely take a very long time for them to investigate each and every single one of them. They’d have to read through the entire email and verify every single statement that she made. By hand, they election would have been over by the time they had completed the investigation.

Machine learning is the perfect technology for this application. We can’t write a program to find sensitive information since we don’t know exactly what we’re looking for. With machine learning, the FBI would have instead written instructions to scan the email, identify all of the statements made, and compare them against a database of what information would be classified. At first, these programs can be quite slow. But every success and every failure allow the computer to change its algorithm. Within 6 days, a computer had completed something that would have taken people months.

Go To Homepage

Popular in the Community