In partnership with researchers at various Asian institutions, Microsoft engineers have put forward a novel method for improving upon the page rankings generated by today’s search engine requests. The new method ‘BrowseRank’ weighs how people actually use the Internet, thereby adding a human factor to the process.

The paper’s authors noted that the importance of a page is likely to be identified by the number of visits by the users to that particular page and the time periods spent by them on the same page. This means more the visits and longer the time period, higher is the importance of the page.

Google’s PageRank measures the relative significance of web pages through the use of a link analysis algorithm (a sequence of data-processing instructions), that allocates a numerical weighting to every element within a given set of hyperlinked documents.

The internet search pioneer stressed that the pages that it believes are significant receive a higher PageRank along with a better chance to feature at the top of the search results. Google also supported its process and technology saying that it has always considered a pragmatic approach to enhance search quality and build useful products, and that its technology employs the collective intelligence of the Web so as to ascertain a page’s importance.

As gauging the relevance of Internet searches enables the search engine majors such as Google, Yahoo and Microsoft to target their ad placements on behalf of clients, it becomes absolutely essential for them (search engines) to adopt this method. However, the software maker along with its collaborators challenged PageRank saying that it misses the mark due to allowing the importance of pages to become artificially inflated. They also claimed BrowseRank is superior to PageRank because it is used on a user-browsing graph that is created from data which reflects actual human behavior on the Internet.

The Internet browsers at Web clients can record the data based on user-behavior, which can be collected at a Web server. BrowseRank’s user-browsing graph can represent the Internet surfer’s random walk process more precisely, which is hence more helpful in calculating page importance, according to the collaborators. In addition, this new method also includes the amount of time spent by users on the pages.

Good news for Microsoft is that the experimental results indicate that BrowseRank certainly surpasses the baseline methods like PageRank and TrustRank in many tasks.

Reacting to Microsoft’s claim, Google said that PageRank is not the only way it uses to rank search results, but relies on above 200 various signals to examine the overall link structure of the Web and decide which pages are most significant. Next, the company undertakes hypertext-matching analysis to determine the pages that are appropriate to the specific search being performed. Finally, the combination of overall importance and query-search relevance allows it to place the most reliable and relevant results on the top.