The Ranking Algorithm of Google Scholar
Link analysis is an important tool used by Web search engines, while it also plays a part in the study of citations among scientific papers and journals. The standard measure of citation impact, impact factor, is the average number of citations received by a paper in the given journal over the past two years. Here, the average attention paid by the academic community to papers published in the journal serves the role of voting by in-links. Google Scholar, however, combines the traits of Web search engine and citation analysis, and is different from both.
Google’s ranking function for web search is believed to involve PageRank as well as non-PageRank methods developed on hubs and authorities, while Google Scholar, like citation analysis, deals with journal articles that have with no in-links or out-links, rather than webs. Citations play the same role in-links play in link analysis, and do so in a more explicit way. They directly show what other articles have influenced one article, and citation counts can be easily used to show how many times one article has been endorsed. As a research by Beel and Gipp found out, the citation counts of an article is the highest ranking factor in the ranking of articles in Google Scholar. Also, the occurrence of a search term in an article’s title has a strong impact on the article’s ranking, while the impact of search term frequencies in an article’s full text is weak. On the other hand, Google Scholar does not seem to use the principle of repeated improvement, as citations from high-impact journals are count in the same way as citations from low-impact ones.
In the same way as Web-page authors would try to understand ranking functions of Web search engine to make their pages rank higher, scholars may also utilize knowledge of the ranking algorithm of Google Scholar to make their article appear the first. And since the factors in ranking algorithm of Google Scholar are more controllable than those in normal web search, scholars may do so simply by using relevant words in the title as much as possible, and increase self-citations or citation alliances. On the other hand, the reliability of Google Scholar as an information source is still debated, and researchers would use the same knowledge to determine the usefulness of articles found on Google Scholar. Thus, there will be a balance in the end between costs and benefits of making paper rank higher on Google Scholar.
Jöran Beel and Bela Gipp. “Google Scholar‘s Ranking Algorithm: An Introductory Overview”. Proceedings of the 12th International Conference on Scientometrics and Informetrics (2009), volume 1, 230–241. http://www.sciplore.org/publications/2009-Google_Scholar%27s_Ranking_Algorithm_–_An_Introductory_Overview_–_preprint.pdf