Open Access News

News from the open access movement


Thursday, April 13, 2006

Comparing with MAS with Google Scholar and OAIster

Charles W. Bailey, Jr., A Simple Search Hit Comparison for Google Scholar, OAIster, and Windows Live Academic Search, Digital Koans, April 13, 2006. Excerpt:

Given that Windows Live Academic Search’s content is limited to computer science, electrical engineering, and physics journals and conferences, a direct comparison of it with other search engines is somewhat difficult.  Although its limitations should be clearly recognized, the following simple experiment in comparing the number of hits for Google Scholar, OAIster (a search engine that indexes open access literature, such as e-prints), and Windows Live Academic Search may help to shed some light on their differences. (Note that OAIster does not typically include content directly provided by commercial publishers, although it does include e-prints for a large number articles published in academic journals.)

The search is for: "OAI-PMH" (entered without quotes).

"OAI-PMH" being, of course, the Open Archives Initiative Protocol for Metadata Harvesting. This is a highly specific search, where many, but not all, hits should fall within the subjects covered by Windows Live Academic Search....To get a better feel for the baseline published literature about OAI-PMH, let’s first do some searching for that term in specialized commercial databases.

  • ACM Digital Library (description): 51 hits.
  • Engineering Village 2 (description): 66 hits.
  • Information Science & Technology Abstracts (description): 36 hits.
  • Library Literature & Information Science Index/Full Text (description): 13 hits.

Now, the search engines in question (the links for the below search engine names are for the search, not the search engine):

So, what have we learned? Windows Live Academic Search has a somewhat higher number of hits than the selected commercial databases and, if adjusted downward for publisher versions only (see below), is on the high end. This suggests that it covers the toll-based published literature well. However, it has a significantly lower number of hits than OAIster and Google Scholar, suggesting that its coverage of open access literature may be weaker than Google Scholar and it is quite likely weaker than OAIster....Of course, this simple experiment tells us nothing about the presence of duplicate entries for the same work in search result sets, which could be important for a meaningful open access comparison....