Skip to main content



PageRank and New Search Metrics

Web search has advanced significantly since its creation, and continues to advance, with improved search algorithms, and new metrics to base search rankings on. The search at Google was initially based on PageRank, an algorithm which assigns each webpage a measure of importance, and while PageRank is still a component of search at google, there are many other components as well. One feature which is changing google’s search is the knowledge graph, which aims to retrieve information for the user and display it so there is no need to click on any of the links provided. Search algorithms are advancing more in language processing, so understanding the semantics of a search query is becoming a more crucial part in search than PageRank. The article notes the results of using these new algorithms, but doesn’t include how the new search features are implemented with old metrics, including PageRank. As an example semantic search, could return a set of pages which could be further ranked using PageRank.

The article notes how the use of display boxes reduces traffic to linked websites, as simple information is aggregated on google’s website. This also would affect advertising, as the click through rate of advertisements could decrease for keywords which now most relevant information displayed on google’s website. The value advertisers have for those keywords will decrease, so on searches with display boxes, the advertising slots would become cheaper. Google has taken this into account, and tailors ads more specifically for the display box being presented, so the ad slots are directly related to the content. An example of this is searching for a musician includes ads in the display box to music streaming services. The method of selling these slots is most likely similar to the auction for keyword slots, but have a much greater value, and google may be selective in who can bid for the slots.

Web search has advanced past using PageRank as the sole algorithm for ranking pages, and is advancing to include metrics such as semantic search. Google is also aiming to deliver information in a simpler way with display boxes, supplanting search results for certain basic queries. Even with these new features of their overall search algorithm, PageRank still plays an important role in search, and is used in conjunction with new features.

Source: http://www.forbes.com/sites/jaysondemers/2014/10/28/what-googles-knowledge-graph-means-for-the-future-of-search/

Comments

Leave a Reply

Blogging Calendar

October 2014
M T W T F S S
 12345
6789101112
13141516171819
20212223242526
2728293031  

Archives