Forum Moderators: Robert Charlton & goodroi
What patents should be the focus (maybe 5-10 of the most important)? Thanks in advance for any feedback!
Just because Google can, doesn't mean they do. And even if they do, a patent does not tell you anything about the implementation.
Try building a space shuttle on the strength of patent info, and you'll never have to worry about re-entry. Because you'll never leave the ground :)
Yahoo! holds patents on PPC, no real patents on most of their infrastructure as it runs on open source software. Their brilliance isn't in the tech, its in the carefully fostered perception of their tech.
For storage try a search on "lustre" file system.
For OS try "linux".
For parallel processing try "map reduce".
Hardware is commodity (upper end consumer processors, disks, motherboards).
No patents available for lots of luck and terrific timing.
patent probing is a pure geek activity
I can't deal with them online, so I've got tons of them printed out and actually do read them. I'd suggest printing out the ones you want to get into, it's definitely a MEGO activity (My Eyes Glaze Over). I find they takes more than one reading - and keeping pencils and pens handy to underline and take notes in the side columns.
The first thing is to scan through the whole thing so you know which parts get mathematical so you can skip over them - unless you're up to the math, which I'm not. What they're trying for is usually condensed into the abstract, so you have to translate that into understandable wording for yourself - and the very act of re-writing it in simple terms helps clarify it.
The claims are hideously repetitious, but read through them anyway and note which of them they keep referring back to - and mark those. It probably won't be clear at that point, but it gives an idea where they're going.
The description toward the end (after the equation stuff, most of which I have to skip) is the part that'll take going over a couple of times, and once those side-column notes are taken and key parts underlined, it's easier to grasp the concepts and be able to go over the major points.
Just the activity of taking the notes while reading helps to clarify, because it'll usually be factors that express concepts - and what you're mainly going after are the concepts rather than the tough mathematical specifics. You'll find by the second time through that you'll spot specific elements that you can recognize, and while it may not seem so, some of them can be directly applied - and comprehended.
Even though just getting or applying for a patent or publishing a paper doesn't mean they actually "use" the thing, there are generally sound elements to be found in the details that are applicable and useful anyway.
it's definitely a MEGO activity (My Eyes Glaze Over)
perhaps I should have said 'nerd' not 'geek'
Joking, honest! ;)
I thank you for making the difficult seem possible ;) I think I will definitely follow your advice (I need to print things out as well!). With over hundreds of patents, which 5 would you say is the most important (or at least had or have the most impact on the search landscape)?
If you were teaching a 3-hour class to patent lawyers (with no SEO experience) on Search Technology and Google patents, which ones would you cover?
- Information retrieval based on historical data
- Detecting spam documents in a phrase based information retrieval system
- Phrase-based searching in an information retrieval system
- Link based clustering of hyperlinked documents
- Methods and apparatus for estimating similarity (Duplicate content)
