Google starts ranking journals

 

In a blog posted on April Fool's Day, Google announced a new feature to its Scholar service. This was no prank. It was the genuine debut of a new tool called Google Scholar Metrics. The service follows the same principle that has made Google's web search engine so successful - when you are unsure what a user is looking for, give them a list of options ranked by a metric of popularity. In this instance, the users are academics ready to submit their next breakthrough but are uncertain which journal to choose. The solution Scholar Metrics offers is a database summarizing the sway of the distributors of scholarship "to help authors as they consider where to publish their new research".
Here's how it works. Google creates a list of all the articles a journal has published in a specified period of time. The citations to each article are counted in order to determine the publication's h-index, which is the largest number "h" such that each of the set of "h" articles were cited "h" or more times. As an example of how the h-index is calculated, consider a publication that has had six total articles having 2, 18, 11, 3, 22, and 9 citations, respectively. This gives the journal an h-index of four. Articles meeting the h-index criterion constitute the h-core. In the example, the core is the articles with 18, 11, 22 and 9 citations. Within the h-core, the median of the citation counts is used to assess the typical influence among the most highly cited set and is reported as the h-median. In the example, the h-median is 14.5.
You might think the "h" of the h-measures was for Helder Suzuki, the Google software engineer who made the blog post unveiling Scholar Metrics. Actually, the "h" refers to Professor Jorge E. Hirsch, a physicist at the University of California, San Diego. Professor Hirsch originally proposed the h-index in 2005 as a means for individual scientists - theoretical physicists, say - to gauge the quality of their work. Hirsch thought total publications, the standard evaluation measure at the time, was a flawed yardstick of a scientist's contributions. In his view, one had to also consider the distribution of citations resulting from a researcher's papers to really know whether that scientist's output was influencing his or her field. And now Scholar Metrics is applying the same idea to the scientific journal.
Scholar Metrics isn't the first tool for evaluating the influence of scholarly publications. In January 2009 Thomas Reuters introduced the impact factor. It is calculated by taking the total number of citations a journal has received in the past year and dividing by the total number of articles it has published in the previous two years. Impact factors are released as part of the annual Journal Citation Reports (JCR). Unlike Scholar Metrics, the JCR is a proprietary service.
So which is the better measure of a publication's influence, h-index or impact factor? If we look at the top ten distributors by each measure, we find Nature, The New England Journal of Medicine, and Science, the major players we would expect to be at the top of the list, taking the 1, 2, and 3 positions by h-index (Figure below). Not so for 5-year impact factor, which doesn't even include Science in its top 10 or PNAS in its top 100. In their place are a number of publications specializing in reviews, which makes the impact factor seem peculiarly out of touch. Review articles, although much read and frequently cited, are not the papers that shape their fields, and these specialty journals wouldn't even be considered by authors debating where to report their next breakthrough.

Further details read source webpage:-http://www.significancemagazine.org/details/webexclusive/1788885/Google-starts-ranking-journals.html