Cut through the clutter.

Home in on key papers, citations, and results.

Leverage AI To Combat Information Overload

With millions of research papers published every year, there is a huge information overload in scientific literature search. Semantic Scholar leverages our AI expertise to help researchers find the most relevant information efficiently. We utilize methods from data mining, natural-language processing, and computer vision to create powerful new search and discovery experiences. Starting with Computer Science in 2015, we plan to scale the service to additional scientific areas over the next few years in support of AI2’s mission of "AI for the Common Good".

Project features currently under development are:
  • Ability to provide an overview or quickly find the most relevant survey papers for a topic.
  • Filtering of search results using automatically generated facets like authors and venues.
  • Identifying "key" citations to overcome citation overload.
  • Extracting and making figures and captions more easily accessible.
  • Identifying and presenting useful concepts and their relationships.

Cite-o-matic: Automated Literature Review

Cite-o-matic is a deep learning model for literature review, specifically trained to learn to give meaningful predictions, even when it’s wrong. Cite-o-matic requires only a title and abstract to give useful results, allowing it to be used at any stage in the writing process.

Learn More

How Semantic Scholar Works

PDF Extraction

State of the art PDF extraction mechanisms specifically targeted to scholarly articles.

Targeted Search Index

For serving relevant results for queries specific to the academic domain.

Customized GUI

A user interface tailored to academic search with features that support the academic community.

“Impossible is not a fact. It’s an opinion.”

Other Projects

  • Aristo

    Answering science questions

    Learn More
  • Plato

    Extracting knowledge from images, diagrams, and videos

    Learn More
  • Euclid

    Solving math and geometry problems