Signals and classifiers: Google ranking secrets

[Italian abstract: un importante sguardo su Google, dall'interno, e su come viene definito l'algoritmo di ranking. Sto scrivendo Ho scritto un'esposizione più completa in italiano è sul blog colletivo Beanbol]

Very interesting article on the New York Times: Google Keeps Tweaking Its Search Engine, it will be quoted lot of times in next hours.

Saul Hansell spent some time with Amit Singhal, the master of the ranking algorithm of Google.

Hansell points out that Google gives a lot of good answers. At the same time, lot of users

click away from Google, disappointed that they couldn’t find the hotel, the recipe or the background of that hot guy. Google often finds what users want, but it doesn’t always.

So, there are lot of little changes on the ranking algorithm. We already know about intrinsic (on-page) and extrinsic (off-page) critera, the most famous is link popularity given by the PageRank. Mr. Singhal told the NYT that there are more than 200 signals. The algorithm then uses classifiers to categorize queries (product, information, company, person,…). The topicality is “a measure of how the topic of a page relates to the broad category of the user’s query”. A final check controls the diversity of results in the results page to ensure that results cover different perspectives. Special explanation are given on QDF, the query deserves freshness: Google looks at what are the hot topic on blogs and gives more fresh results on these.

Some comments by other bloggers.

SEOmoz has a very important comment. The ranking algorithm is different for some queries:

The short paragraph about the French Revolution, if accurate, gives some insight into the fact that the algorithm is not uniform – not even close. Individual queries get individual attention

Furthermore, he notes that these are

a lot of confirmation about what many have only theorized until now.

Data Mining has some doubts on the quality of research: the article speaks about tweaking Google, that sounds not really serious. Probably it is just a journalistic language. He points out that everything is based on ranking. I think that maybe we will see in next years the end of ranking based on one criteria.

Google Blogoscoped makes a summary of most important concepts: users expect more, balancing changes, how search problems are escalated, ensuring fresh pages don’t get lost, ranking pages through signals, classifiers and topicality, searchers often use ambiguous queries.

via Slashdot

Why do I blog this (©NN)?
Google’s secrets, this is very interesting for my PhD research, for my web promotion activities and for a general knowledge of online communication.


About Marco

Marco Faré, MSc, online communication specialist (Lugano, Switzerland)
This entry was posted in net & web, phd research and tagged , . Bookmark the permalink.

One Response to Signals and classifiers: Google ranking secrets

  1. Pingback: www.BeanBol.com » I segreti del ranking di Google

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>