Analytics Dispatch 12/16/2019
Instagram’s recommender system
MORE LIKE THIS
Powered by AI: Instagram’s Explore Recommender System
Recommending the most relevant content out of billions of options in real time at scale introduces a ton of machine learning engineering problems. Here’s a detailed look at the system Instagram built to provide its users with personalized content. - Instagram Engineering
PLAY FOR KEEPS
What Board Games Teach Us About Data Visualization
“[A]t their core, all board games are data visualizations. Data and information are visualized as pieces of different colors and shapes (called meeples) placed on boards specifying coordinate systems. The rules of the game determine how the current situation of the data can be transformed into a more desirable state.” - Nightingale
ALL LANGUAGES ALLOWED
dbplyr: A Path to More Inclusive Data Transformations at the ACLU
By using the Tidyverse’s dplyr to generate complex SQL code in their data warehouse, the ACLU analytics team is able to create reusable data workflows that everyone can benefit from—whether they’re using SQL, Python, R, or a BI tool. - ACLU Tech & Analytics
STRING THEORY
The User-Agent — That Crazy String Underpinning a Bunch of Analytics
How did this one piece of data become so vital to pretty much all web analytics? How is it used? And what caveats does it come with? - Randy Au
STUDY UP
Supervised Machine Learning Case Studies in R
This free course covers exploratory data analysis, preparing data so it’s ready for predictive modeling, training supervised machine learning models, and evaluating those models—all using real-world data. - Julia Silge