In The News: AI for Better Dating, Hospital Operations, and Censorship
And we’re back with another weekly newsletter! This time, we tried to dive into some applications of data science that are super close to our daily lives, but rarely at the top of our minds.
People Are Dating All Wrong, According to Data Science
by Spencer Karp
These days it seems like data science has an answer for all the issues in our lives. It can predict our spending habits and understand our routines, so can it be the answer to our miserable dating lives? According to this article: maybe… Much of the current literature on data science and dating is limited because the sample sizes in these studies haven’t been large enough to truly be conclusive. However, Samantha Joel has solved this problem. By pooling together data from many different studies, she is able to get a clearer picture of what data tells us about the ideal partner. This is a perfect case study if you are looking to find data to answer a question when it seems like there is no data. So what did she find? Well, she found that the ideal partner does not exist. Most of the characteristics that we normally attribute to compatible partners (e.g. looks, wealth, interests) don’t actually have any correlation with compatibility. One thing data science can predict is how we view compatibility, but not what partners are actually compatible with us. Read more about Joel’s findings and how we can be better partners here.
The Unnerving Rise of Video Games that Spy on You
I grew up playing video games, multiplayer or single-player, RPG or FPS, you name it. But I don’t immediately think of it in terms of data, especially my data– “The user, by acting in ways that comply with the rules of the game and the specifications of the hardware, is parsed as data by the video game.” One 2012 study explored World of Warcraft data detailing the tasks a player completed in-game (number of deaths, fish caught, etc.). In the study, the researchers surveyed the players for personality traits like conscientiousness and found a correlation between risk-taking or brash game actions like falling to death and lower self-reported conscientiousness in real life. On the other hand, those who felt they were more conscientious devoted more time to mundane tasks in-game (fishing was the example used). With this level of insight possible from player data, it’s no wonder that major game studios like EA and Activision have disclaimers of data collection in their license agreements. The business applications of understanding player behavior are clear, from knowing how to induce more spending on in-game currency to understanding how to drum up certain emotions to enhance story quality and sell more copies. As much as I enjoy both video games and data, this casts a dark shadow on the industry. Gaming companies can take advantage of addictive tendencies, or, in the case of Tencent complying with China’s efforts to curb video game addiction, firms can even employ techniques like facial recognition to surveil and restrict players. Just some food for thought, the next time you decide to Leeroy Jenkins into a boss battle.
Predictive Analytics Streamline Hospital Operations at Seattle Children’s
by Annika Lin
Seattle Children’s Hospital leverages a predictive analytics strategy to test new processes to streamline hospital operations and save resources. Their digital twin simulations strategy creates a clone of the hospital for data scientists to experiment with potential improvements. For example, determining if adding an admission boarding area in the emergency department will boost operational efficiencies. By using years of patient and hospital utilization data, the children’s hospital can also investigate questions like how rapidly the hospital can be filled again once it restarts elective surgeries. Discrete event simulation also helps predict a capacity action score based on daily bed capacity and labor resources. Because failing to provide adequate resources is expensive, this predictive capacity has significantly increased efficiencies and savings. As events like the COVID-19 pandemic have tightened hospital resource constraints, tools like predictive analytics are essential to optimizing healthcare operations and costs.
