Skip to main content

Research at Twitter Cortex

Applied Machine Learning Research to fuel the public conversation

Abstract nodes and lines on top of a red background texture.

Areas of Research

The Applied Research teams at Twitter Cortex focus on eight key areas of exploration: Content Understanding, Learning Methods, Recommender Systems, User Modeling, Machine Learning Ethics, HITL, Graph Learning, ML Ethics and Transparency Research, and Data Science.

Content Understanding

Using Machine Learning to understand content to classify, filter, and improve recommendation performance.

See publications

User Modeling

Use machine learning to understand the behavior of users (e.g. what they’re interested in and how they interact).

See publications

ML Ethics

Research, development of best practices and tools to advance the responsible use of algorithms, proactive risk assessments and mitigations.

See publications

Learning Methods

Leveraging new machine learning paradigms such as self-supervised learning and graph learning within our products, and generally improving how we do machine learning.

Human-In-The-Loop

Twitter data is public by design. HITL is an inherent part of the product lifecycle, decision-making, policy, and user research.

Data Science

Twitter Data Science embraces a culture of innovation in how we utilize the rich data from our platform, resulting in a direct impact on our core company objectives.

Recommender Systems

Improving the performance of the systems that recommend content and ads to our customers.

See publications

Graph Learning

Graph structured data naturally exists in many forms and at a scale from thousands to billions of nodes, leveraged for many product areas.

See publications

Other Publications

Additional applied research and publications that don’t fall into any of the above categories and topics can be found here.

See publications