I own Registered Patent #7,082,470 [patft.uspto.gov] in a similar space so I wanted to comment on what I see in this application. The application appears to be written by attorneys and not experts with knowledge and history of link exchange, reciprocal linking, and SEO.
As near as I can understand this it contains good news and bad news. The bad news starts with the title "Identifying excessively reciprocal links among web entities" which implies all reciprocal links are somehow bad.
The negatives continue until about the last third of the doc where they admit that relevant links are good and legitimate and talk about programming an algorithm to specifically identify good linking pages.
From my perspective I would say the most positive thing here is that if webmasters follow the same recommendations I have been preaching here at WebmasterWorld for the past few years regarding the practice of obtaining relevant links through reciprocation in slow/natural volume, they will meet Yahoo!'s positive "degree of confidence" requirements based on the statement that "the closer the value of an entity's function of interest is to the specified threshold for that function of interest, the less uncertainty there is that the value has been artificially inflated."
Also, the emphasis on punishing hosts and domains that specialize in helping spammers implement their schemes is a good argument against all the full duplex automated systems offering irrelevant links in high volume through reciprocation, likewise the fact that they have an alleged way to hunt out and punish sites buying into three- and four-way schemes.
The really bad part is that there's no verbiage validating the essential nature of valid one-way linking and relevant reciprocal linking to the overall functionality of the web. Maybe they do not include this because they don’t want to give away what really works. That has always been Google’s method for writing patent apps.
What this patent application says right at the beginning (almost in a throwaway line) is that Yahoo considers most metatags as spam. Than it goes on to say that Yahoo considers most links spam. It also implies that if you have proper metatags and links you won't be punished for them.
But nothing is said about rewards. Which raises the question, if you don't base part of your ranking decision on things like keywords or links, how do you rank pages?
If you rank them only on, in Yahoo's words, a page's “visual content” you've created a system that is so incredibly easy to spam it's ridiculous. Visit the top four or five pages for your search term, download their content and combine the best parts of each to create your content. All you need to do is change some of the vocabulary and sentence structure to make sure it won't throw up a "duplicate" flag.
One side note, if Yahoo really implements this, their returns are going to start becoming vastly different from Google's since Google still counts positive relevant links obtained through reciprocation.
Such a strategy will probably push Yahoo further into the margins than it already is.
