The purpose of the Eigenfactor algorithm is to estimate the relative influence of reference items based on cross-citation data. Eigenfactor's approach is to rank journals much as Google ranks web pages. While Google uses the network of hyperlinks on the world wide web, Eigenfactor uses citations in the academic literature. Here we use citations within the Microsoft corpus. By this approach, we aim to identity the most influential journals, where a journal is considered to be influential if it is cited often by other influential journals. (See West and Bergstrom for further methodological details.) This iterative ranking scheme accounts for the fact that a single citation from a high-quality journal may be more valuable than multiple citations from peripheral publications.
Like Thomson Scientific's Impact Factor metric, Eigenfactor measures the number of times that articles published during a census period provide citations to papers published during an earlier target window. While Impact Factor has a one year census period and uses the two previous years for the target window, Eigenfactor has a one year census period and uses the five previous years for the target window.
The Eigenfactor® score measures the total influence of a journal on the scholarly literature, or, comparably, the total value provided by all of the articles published in that journal in a year. This is the appropriate metric for making subscription decisions. All else equal, larger journals will have more citations and larger Eigenfactor® scores and will be visited more often by researchers.
If one wants to estimate of the importance of an article by the company it keeps, the size of the journal in which it is published is not relevant. Instead one would want to measure of the average influence of articles appearing in the same journal. The Article Influence® for a journal is proportional to the Eigenfactor divided by the number of articles. This measure is more directly comparable to Thomson-Reuters' Journal Impact Factor. (Note: the Article Influence scores are not shown for this demon version.)
A unique capacity of the Eigenfactor® approach is that journals can be sorted algorithmicly into their component disciplines using the Map Equation (Rosvall and Bergstrom 2008 Proc Nat Acad Sci USA). The Eigenfactor categories obtained in this way form a hard partition in which each journal belongs to only one category.
This approach to mapping the structure of science is empirical: rather than use our preconceived notions about what the structure of clusters or "fields" within science should be, we let the data - in our case, citation patterns - tell us what the clusters or fields are. In other words, we are interested in mapping science according to what researchers do, not what they say that they do or how they self-identify. One interesting consequence of this approach is that the fields vary widely in size according to their citation behavior. Some fields, such as Art Education, are very small and comprise only a few journals; others fields, such as Biology, are very large.