• http://seo-theory.blogspot.com/ Michael Martinez

    There are quite a few problems with this kind of system.

    First of all, reputation has never been a useful mechanism for evaluating quality of content on the Web. However, let’s assume that the mechanism itself is better than we have today.

    What is to keep people from simply copying reputable content from location A to location B (with the digital signature intact)? What would be required to produce digital signatures that validate only at the original point of insertion? What do you do if the site you contributed content to goes through a major redesign?

    Also, if you go with the “seed set” concept, there is an inherent and incurable flaw there. Who picks the seed set of experts, and how do you deal with the inevitable conflicts of opinion and bias that arise within communities of experts?

    Wikipedia cannot even build uniform communities of experts in its growing mob of biased opinions, but once it does get some experts involved they invariably become frustrated with the overwhelming naivity of the Wiki community and leave.

    Various forum packages already offer a reputation system now and it’s mostly a “here’s some rep for my buddy” system.

  • http://www.seobythesea.com Bill Slawski

    All very good points, Michael.

    I do think that they provide some hints of things that they might try to do to address some of those issues, but those aren’t fleshed out in any level of detail.

    For instance, the application of a digital signature to a specific piece of content at a certain address may create meta data that shows where the content is located, and when. If the creator of that content doesn’t want that content syndicated, then other instances of the content might just be ignored by a search engine. Could that be manipulated? Maybe.

    I was left with a number of the same questions that you have, though.

  • http://www.comstockfilms.com/blog/tony Tony Comstock

    What about ideas, transmitted digitally, that do not yeild easily to language, or language based tools? Will this new app deal with them any better than what we have now?

  • http://www.seobythesea.com Bill Slawski

    Hi Tony,

    The patent application does mention that non-text content, such as images, audio, and video can be digitally signed by agents (writers, editers, reviewers, etc.) and associated with those agents. I think that’s what you are asking about.

    We are only given a broad overview with the patent application, so it’s a little hard to tell what might be done with the system described here.

    Could it be used, for instance, to assign greater or lesser amounts of reputation to users who leave annotations regarding specific videos or podcasts or so on? Maybe. Or based upon the reputation of the creator of that video. Probably. Will those reputations be used to influence rankings of pages that content appears upon. It could.

    Will it be used? We don’t know. Possibly.

  • http://www.seopittfall.com pittfall


    Could there be a way to assist this process? Signatures and comments are typically labeled in social websites like blogs, however, what about standard websites? Many of these would be considered experts, and be considered an important aspect of ranking. Are there any considerations for unknown authors?

    Very interesting topic!

  • http://www.seobythesea.com Bill Slawski

    Hi Steve,

    The patent does address what might happen when there are no digital signatures associated with content on pages:

    If no digital signature associates an agent with a specific content item, the content item is associated with an owner of a location where the specific content is found and a score is assigned to the owner based on the specific content item.

    So, a digital signature isn’t absolutely essential to how agent rank might work, though it does look like it is important if information is going to be collected about different authors from the same page.

    If Google does attempt to implement this, it will be interesting to see how they try to inspire people to use signatures.

  • http://andybeard.eu AndyBeard

    You walk into a bookstore and the material available is categorised in various ways, but they also have a top selling authors section, and a recent hits section.

    People have been signing their work in various ways on the internet for a long time. I sign my work with a link back to the original work on my own domain.

    Adoption of various forms of digital signature are becoming more widespread, and for example you can use things like OpenID combined with blogging comments.

    The invention seems too “obvious”

  • http://www.seobythesea.com Bill Slawski

    Great analogy, Andy.

    To simplify the test that is used when considering granting a patent, the work in question must be “new, nonobvious, and useful.”

    I mentioned Ted Nelson’s work because this patent application echoes the idea that snippets of pages from other sources and other authors can be joined together in new documents, with references to the original sources and links to those that don’t go away and are trackable.

    Digital signatures, though not as widely used as they could be, through initiatives like OpenID also aren’t anything new.

    As I mentioned in my conclusion though, the answer that appeared to elude Ted Nelson, who has been working on something like this for decades, has been a way of tracking and ranking those content items across different pages and sites, and indexing pages based upon a reputation system.

    Is that aspect of this invention “new, nonobvious, and useful?” That’s going to be the task of the patent examiner to determine. That doesn’t mean that we can’t question it, like you have. But I do think that there are aspects of combining these many ideas together that are nonobvious even though parts of them may be.

  • http://www.ginside.com/ Jonathan

    Although I’m reading the actual patent process and I’m not seeing anywhere that it claims it’s owned by Google. The inventors are yes, up in Mountain View, but where does it actually state it’s for Google?

  • http://www.seobythesea.com Bill Slawski

    Hi Jonathan,

    It says that in the USPTO assignment database, which records assignments of patents and patent applications.