Google Patents - how they can help you better understand Google's rankings
Do you want to know how Google search engine works?
How they rank web pages?
How they crawl websites, measure and calculate things like domain information, user behavior, website trust, authority, link weight, freshness of links etc...?
Or do you want to know how other things works, like near duplicate or duplicate content, or image search or code search or how they rank web pages locally...
OK, I know, you get it, so lets move on...
Now, you know that when you have a very good idea and you want to protect your idea from others stealing it from you, you need to patent it, in order to patent your idea you need to carefully document your invention, make a working model and so on...
You see, Google have employed some great minds there and they have made some great inventions and projects, so in order for them to keep the right on their inventions, Google needs to patent them and now you are guessing where this is going to ;).
You don't even need to go to some .gov websites or some offline facilities and search thousands patents, Google have made that easy for you too :).
Google have search engine just for patents and if you go there and search in quotes
"Assignee: Google Inc" you will find all patents made by Google. This is the link for that query search: patent search.
Most interesting Google patents for you would be documents(web page) indexing and ranking patents, like these (pdf files are in better resolution):
Information Retrieval Based on Historical Data
Search result ranking based on trust
Ranking documents based on user behavior and/or feature data
+ you can find others also and connect the dots ;)
Now I know that these documents are written in very "techy language", but read it very carefully and use illustrations(pics) to figure out these documents.
Also I know that these are not fully detailed algorithms, but they can give you very good information how things works and if you read and connect few other Google patents you will have great understanding about things behind the closed doors.
There are few other patents that might be interesting to you, like:
Detecting duplicate and near-duplicate files
Techniques for finding related hyperlinked documents using link-based analysis
Link based clustering of hyperlinked documents
Adaptive computation of ranking
and many, many others, just search with this search string "Assignee: Google Inc" and find the one you are interested in.
Back on the Google Ranking Algorithm, yes this is very tedious task to go trough all over these patents/documents, but this is the best way to find great and valuable info., just few things from these documents can make very big difference for your business.
As I said it before these are very "techy" documents, but luckily for you, guys from SEOmoz.org "translated and rewritten" these in non-techy language, so regular person can understand what is happening, but again I recommend that you read original document also, you will get better and little more detailed picture.
So here is the link from SEOMoz.org with simplified version of Google Indexing and Ranking Patents:
SEOmoz | Google's Patent - Information Retrieval Based on Historical Data
This is it for now, hope you like it and find it useful,
Nemanja