Google has secrets. Despite their promotion of open source, open spectrum, open mobile, and overall openness, they are not very open about their algorithms. And who can blame them? They have a business to run.
Manber said the heart of the group is the core ranking team. While PageRank is the “most famous” part of Google’s algorithm, Manber pointed out that other factors work into the system, including:
- Language models (phrases, synonyms, diacritics, spelling mistakes, etc.)
- Query models (how people use language)
- Time models
- Personalized models
Next up is the evaluation team. Did you know….
- There are automated evaluations every minute
- There were over 450 new algorithm improvements in 2007 (that’s about 9 per week)
Manber says that most of the improvements are related to relevancy, but some projects are dedicated solely to simplifying algorithms.
Developing new features and new user interfaces is the responsibility of yet another team. Last year’s Universal Search update is attributed to this team as well as Google Notebook, Custom Search Engines, and many of iGoogle‘s improvements. This team is so dedicated to users, that it sometimes goes into people’s homes and observes them as they search!
Of course, there is a spam-fighting team. Black hats beware, this team focuses on hidden text, off-topic pages, keyword-stuffed pages, and other ways people attempt to abuse rankings. The team is in cohorts with the team at Google Webmaster Central, which gives them access to feedback from site owners.
Manber said there are other teams working on specific projects, but didn’t get into any additional details – yet. He said to look for more posts in the future that will inform about updates to search quality.
What do you think about Google’s “transparency”? Did they really offer up anything new? Tell us what you think by leaving a comment.