Searching on Cobaltmetrics

Cobaltmetrics can only be queried by identifiers. Our team has a strong background in natural language processing and hands-on experience with bibliometrics. We know how ambiguous natural language can be, especially named entities like the names of persons and organizations. We decided early on to focus on speed and relevance, so searching by names will most likely never be supported. And if you are one of the 47,000 people in the US named John Smith, please register for an ORCID identifier.

Query expansion

All identifiers in Cobaltmetrics are stored as URIs, so every search term will consist of a scheme, a colon, and a value. For example, to search for RFC 3986, enter rfc:3986. Because all URLs are URIs, you can also enter full URLs into the search bar, and we'll do our best to extract identifiers from them, for example https://tools.ietf.org/html/rfc3986 instead of rfc:3986. The statistics API lists all schemes (i.e. identifier types) that are currently supported.

Our search engine will automatically normalize identifiers depending on their type, for example if they are not sensitive to case like DOIs, or if their canonical form includes hyphens like ISSNs or ORCID iDs. Likewise, our search engine will reject syntactically incorrect identifiers.

By default, Cobaltmetrics will search for the identifiers you entered and, through a process known as query expansion, all other identifiers known to identify the same document or digital object. For example, we know from PubMed that pmid:23193287, pmcid:3531190, and doi:10.1093/NAR/GKS1195 refer to the same article. Query expansion allows you to find all citations and backlinks to a document with multiple identifiers, with a single query.

See our page on URI transmutation for more details.