FAQ Question
How are documents ranked?
Semantic Scholar uses a relevance function for ranking documents that takes into account different aspects of the query and the paper details to provide the most relevant results.
You can sort your search results by relevance, citation count, most influential papers, and recency using the dropdown menu in the upper-right on the search results page.
The complete Semantic Scholar search reranker model is available at https://github.com/allenai/s2search. You can also read "Building a Better Search Engine for Semantic Scholar" on the AI2 blog.