Internet search engines, which make information nearly universally accessible, still perform poorly at handling queries with mathematical content. The reasons for suboptimal handling of mathematical content by search engines include the following: the difficulty of entering mathematical expressions into search engines, the lack of standard encoding for mathematical content, the difficulty of determining if a given page or information source is relevant to a mathematical query, etc.

This disclosure describes techniques that address the aforementioned problems, e.g., by enabling easy entry of mathematical expressions into search engines, by creating a standard representation of mathematical content based on a machine-learning embedding of mathematical expressions, and by identifying pages relevant to a mathematical query using a machine-learned ranking system.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.