Data science starts with a question and ends in insight.

Data science is a field that draws on tools from computer science, statistics, and mathematics to turn raw, unstructured information into knowledge we can act on. 

At Columbia, we take a broad view of data science—one that builds on core technical skills to encompass the full data journey, from formulating questions to applying insights in real-world contexts.

This begins with defining problems, collecting and cleaning data, analyzing it, interpreting results, and applying insights to address real-world needs.

Columbia data scientists learn to understand context, collaborate across disciplines, and communicate findings clearly and responsibly. 

These are both the essential skills that distinguish sought-after leaders from strong technical contributors, and they are also foundational to our commitment to Data for Good: using data science with purpose, integrity, and awareness of its impact.

We often discuss data science and artificial intelligence (AI) together. That’s because while data science is a transformational tool in its own right, it also underpins AI. 

AI systems depend on data science to prepare, interpret, and validate the data that powers learning and automation. At Columbia, this relationship is central to how we approach both fields—applying them together to address real-world challenges with ethical and societal awareness.

Defining Data Science

The definition of data science rests on three foundational concepts: study, extraction, and value.

Study

More than just analysis, study reflects the blend of scientific rigor and creative inquiry that guides this dynamic field. Data science is profoundly applicable across diverse domains.

Extraction

This is the active, intricate process of transforming raw, often messy inputs into usable knowledge through tools like computation, modeling, and visualization.

Value

Value depends on context. Whether it’s informing policy for a government, advancing discovery in medicine, or driving revenue for a corporation, value is the benefit drawn from the process.