Data Science

Gaining insights to deliver meaningful social interactions

Data scientists at Facebook conduct large-scale, global, quantitative research to gain deeper insights into how people interact with each other and the world around them.

Our findings directly inform decisions to improve people’s everyday experiences on Facebook, make it easier and more intuitive to use, and find ways to facilitate meaningful social interactions. Research efforts span a variety of disciplines, including computational social science, econometrics, operations research, market intelligence, survey science, and statistical computing. We employ a mixture of methods to accomplish our goals, including machine learning, field experiments, surveys, and information visualization. We also build scalable platforms for the collection, management, and analysis of data, and actively contribute our scientific findings to the academic research community.

For information about Core Data Science research at Facebook, visit the Core Data Science page.

Latest Publications

All Publications

PLOS ONE Journal - September 8, 2021

Large-scale decrease in the social salience of climate change during the COVID-19 pandemic

Brian R. Spisak, Bogdan State, Ingrid van de Leemput, Marten Scheffer, Yuwei Liu

AISTATS - August 31, 2021

Causal Autoregressive Flows

Ilyes Khemakhem, Ricardo P. Monti, Robert Leech, Aapo Hyvärinen

SIGKDD - August 19, 2021

Network Experimentation at Scale

Brian Karrer, Liang Shi, Monica Bhole, Matt Goldman, Tyrone Palmer, Charlie Gelman, Mikael Konutgan, Feng Sun

SIGKDD - August 14, 2021

Clockwork: a Delay-Based Global Scheduling Framework for More Consistent Landing Times in the Data Warehouse

Martin Valdez-Vivas, Varun Sharma, Nick Stanisha, Shan Li, Luo Mi, Wei Jiang, Alex Kalinin, Josh Metzler

Explore Data Science

 

Automatic Alt-Text (AAT) allows screen reader users the ability to understand the content of most images (hopefully all images soon!) in News Feed.

 

The reach of Facebook friendships across European cities.

 

Calculating three and a half degrees of separation.

 

Researchers explored the dynamic of jobs running in families. This network visualization shows how much more likely a child of a parent in one profession is to choose another profession vs. someone from the general population.

 

Example social network showing exposure to diverse information on Facebook

 

Visualization of popular Indian politicians and political organizations for each state and their fan following across states. Visualization inspired by mbostock’s hierarchical edge bundling.

 

Figure shows different cities in India and the most prominent emotion among the posts from that city as they celebrated Independence Day.

Core Data Science at Facebook

3:09 | April 23, 2020

Open Source Projects

1:36 | August 16, 2019

Social Connectedness Index

1:11 | July 1, 2019

Social Good Projects

1:24 | July 1, 2019

Careers at Facebook Research
Want to solve some of the most challenging technology problems?