Find Similar Papers

mosaic similar lists papers related to one you already know, from its DOI or arXiv ID: no query to write. Use it to grow a reading list around a key paper.

Usage

mosaic similar 10.1063/1.4972269
mosaic similar arxiv:2304.06563 --max 20 --sort citations
mosaic similar 10.1063/1.4972269 --oa-only --download
mosaic similar 10.1063/1.4972269 --output related.bib

The identifier is a bare DOI, a DOI with a doi: or DOI: prefix, or an arXiv ID with an arxiv: or ARXIV: prefix. The results take the options of mosaic search: --oa-only, --pdf-only, --sort, --output, --download, --zotero, --obsidian, --json.

How it works

Source	What it returns	Needs
OpenAlex	the paper's related_works, a list OpenAlex computes for each work	nothing: always queried
Semantic Scholar	its recommendations for the paper	a Semantic Scholar API key

mosaic looks the paper up on OpenAlex, takes the first --max (default 10) of its related works and fetches their metadata; the table's title line names the paper. With a Semantic Scholar key configured (mosaic config --ss-key KEY, free at semanticscholar.org), it also asks Semantic Scholar for up to --max recommendations and merges the two lists by uid, keeping the richer record. A paper that OpenAlex does not know is reported as Paper not found and the command exits with status 1.

--max is an upper bound: OpenAlex lists a limited number of related works per paper, and a related work it cannot return is dropped. --sort relevance ranks the results against the seed paper's title.

The results are saved to the cache like those of a search, so mosaic search --cached and mosaic index see them.

Limits

• Related works reflect OpenAlex's topic model, not the paper's references: expect neighbours in the field, not a citation list. For citations, see the citation graph.
• Without a Semantic Scholar key, the list is OpenAlex's alone.

See also: tutorial: following a field, recipes, mosaic similar.