ReaderMeter: researcher-level metrics based on readership

In a blog post last week, Dario Taraborelli officially announced ReaderMeter. ReaderMeter takes the usage data from reference managers (starting with Mendeley) to analyze the impact of publications by a particular author.

ReaderMeter is a welcome addition to other metrics of researcher impact, most of which are citation-based. And ReaderMeter was hacked together in a few nights, so the service should improve over time. Dario mentions some of the current limitations in his blog post:

  • Usage data only from Mendeley. We don’t know whether the reading patterns of users from other reference managers (e.g. CiteULike, Zotero, Endnote Web and Refworks) differ. Dario is working on adding more services.
  • Author disambiguation. This is a critical component, and the ORCID initiative is probably the best tool to overcome this limitation, especially when data from several reference managers have to be integrated.
  • Article disambiguation. Multiple copies of the same reference are common in reference manager databases. The DOI can solve this problem for those publications that have a DOI, but consequent use of the DOI by journals, citation databases and reference managers has yet to become standard practice.

ReaderMeter is visually pleasing and fun to use. And the data are also provided in JSON format, which allows integration into other web pages, e.g. the researcher’s home page.