If you text even just a few times a day, there’s a good chance you’ve had that moment of clarity when you realize your phone knows you better than you know yourself, as it auto-filled a phrase that you use often, or auto-corrected your common misspelling of a proper term like a band name or movie you regularly reference.
These are examples of your phone using predictive analytics, at a very baseline level. As the name implies, predictive analytics use history-related algorithms and memory to “predict” what will happen in the future. In your phone’s case (no pun intended), it has learned from the history of your texts to predict what you will text in the future.
Here’s how it all works:
The term “analytics” is quite broad, but can be broken down into three categories that help define it, as a whole: descriptive, prescriptive, and the aforementioned predictive.
Descriptive analytics is the collection of data followed by the organization and presentation of what was learned from analyzing the data. Some practical, every day examples of descriptive analytics would be determining the most popular travel destinations of 2019, or, more relative to texting, setting up parental controls on social media to see your child’s screen usage habits.
Prescriptive analytics is utilize both descriptive and predictive analytics to determine the best courses of action for an individual or business. In short, descriptive tells us what happened, predictive tells us what to expect, and prescriptive ties the two together to determine what to do next.
How Does My Phone Know That?
Just as computers need to be programmed before we can use them, our phones need to “learn” about us before they can truly build a good predictive analysis on what we might want to type. The part of your phone that gives you suggestions for quick fills while texting comes programmed with data from the general public. This is why, when you first get a phone, it may frustrate you a lot changing things like “Zakk” to “Zack” even though the former is actually your friend’s name (with the latter being a more common spelling and thus, pre-programmed into your phone based on descriptive analytics regarding the general public).
Every time you re-type “Zakk,” your phone is collecting personal data that will soon outweigh the common data that was pre-programmed, ultimately using that personal data for predictive analysis of your word use. Over time, your phone gets to know you better and better and will be auto-filling the word “coooooool” to your exact number of preferred “o’s.”
How to Train Your Android (Or Any Other Type of Smartphone)
There are ways to “train” your telephone to make this transition from common data to personal data quicker and more effective. Whether pre-loaded on your phone, or a keyboard you’ve downloaded, almost all of them will give at least three options when typing a word the phone does not recognize. The first is simply ignoring the requested change, keeping the word as is, and moving on. This will tell your phone that you prefer the “incorrect” use, but will still view the word as “incorrect” at other, unrelated mentions. If you decide to add the word to your phone’s dictionary, it will no longer view that word as “incorrect” and will most likely not flag it for change again.
If you accept a correction just to move past it, your keyboard will continue to make that correction, and most likely result in you using some choice words to describe it. Simply adding the word to your phone’s dictionary will help keep a more positive relationship with your telephone!