A recent paper on “Finding Scientific Gems with Google” tested the application of “the Google PageRank algorithm to assess the relative importance of all publications in the Physical Review family of journals from 1893—2003″. The results, highlighted on IoP’s physicsweb, indicate that in general highly cited papers also have high Google rank numbers, but the algorithm also found a few exceptional papers that have anomalously high Google rank numbers compared with their citation rank. Such anomalies can occur when several derivative follow-up papers, written at a more accessible level, capture most of the subsequent citations. Google Scholar is implementing similar ranking ideas via the new ‘recent articles’ option, which promises to rank papers based on factors like the number of citations and “the prominence of the author’s and journal’s previous papers”.
The updated Google Scholar conveniently includes links to PDF versions of papers from the authors’ websites (when available), which can be a much faster download option than browsing through the subscription-based publishers’ sites. By contrast, Academic search from Microsoft only links to published or arXiv versions of the papers. Both services, however, are not yet as complete as the traditional citation search databases and, in some cases, have trouble identifying multiple links to the same paper, when respective sources contain typos or use different formats for citations.