The Evolution of Google Ratings (Experts Delegating Authority)?

Imagine Google offering a rating system where you can ask experts, or authorities, to provide ratings for pages on the Web such as scientific articles, tutorials, news stories, or editorials, or products and current movies. It’s something that we might see from Google in the future, if they follow through on a process described in a patent granted to them earlier this week.

How might it work?

You log into this evaluation system and present what you want rated, and you’re provided with a list of primary authorities that you can choose amongst to rate your page or product or movie.

For example, you’re interested in learning about Diabetes, and you run across an article that tells you about how a version of vitamin B1, Benfotiamine, may potentially be useful in fighting off some of the damages that high blood sugar might do to your body. You’d like to know what some experts in the field might think about the article, and you’re shown authorities in the field listed as authorities on the subject, such as the American Medical Association, the Harvard Medical School, and the American Diabetes Association. You select those three organizations to provide ratings.

Those authorities, and people whom they’ve delegated some of their authority to, provide ratings for the article in a manner like you might see in Amazon.com or Slashdot.

The important part of the patent is that this rating system involves experts on specific topics, who might delegate some of their authority to others when providing ratings.

The ability for expert raters to delegate some or all of their authority to provide ratings creates a way for the evaluation system to scale to greater amounts of content.

The American Medical Association might delegate some of their authority to provide ratings to a team of doctors who have applied or have been appointed to the position. The Harvard School of Medicine might include a professor and a clinical team of students to be evaluators. The American Diabetes Association might choose volunteers specializing in Diabetes research to rate medical articles.

If the evaluators research the article I mention above, they might note that one of the authors is the chairman of the Juvenile Diabetes Association, is the head of a research team involving Diabetes at Einstein College of Medicine in New York, has won a number of awards for his research, and that the article has been heavily cited in a positive manner in a number of medical journals since its publication.

The patent is:

Delegated authority evaluation system
Invented by W. Daniel Hillis and Bran Ferren
Assigned to Google
US Patent 7,844,610
Granted November 30, 2010
Filed: May 25, 2004

Abstract

The invention provides an evaluation system for reliably evaluating large amounts of content. The evaluation system is managed by a primary authority that designates one or more contributing authorities by delegating to each a specific quantity of authority.

Each contributing authority may in turn designate and delegate authority to one or more additional contributing authorities, subject to the restriction that the total quantity of authority delegated does not exceed the quantity of authority the contributing authority was itself delegated.

Each contributing authority, and optionally the primary authority itself, may evaluate one or more portions of content by associating a rating with each evaluated portion of content. A composite rating for a particular portion of content may then be determined based upon the ratings associated with the portion of content. Preferably, the ratings are combined in a manner that affords a higher priority to the ratings provided by contributing authorities to which a greater quantity of authority was delegated.

The ratings provided by authorities and their delegates may then be available to others through annotations left with the pages. The patent doesn’t detail how those might appear, but they might make those available in search results or possibly on something like Google’s sidewiki

The ratings may be presented to reflect a number of different senses about the pages or products being evaluated, such as credibility, trustworthiness, accuracy and impartiality.

Primary authorities might be public entities, such as the American Medical Association, or private entities, such as individuals with trusted web presences or peers of the people searching for evaluations.

The identities of the people providing ratings would be known rather than anonymous.

When a primary authority, such as the American Medical Assocation delegates some of its authority, it might start out with a certain amount of authority points, like 200. It might designate 100 of those authority points to a specific doctor, and 10 points each to 10 medical students working in a clinic with that doctor. The doctor could then delegate 50 of those authority points to a co-researcher. If some, or all of those delegates provide ratings for a particular article, the final amount of authority involved in the rating wouldn’t exceed the original amount of authority provided to the American Medical Association. The total amount of authority used in providing the rating would figure into a final rating when multiple sources are selected to provide an overall rating for a page or product or movie.

Conclusion

Will we see a system like this emerge from Google sometime in the future?

I thought it was interesting that the patent notes that experts or “primary authorities” could be private individuals, including peers of the people asking for ratings.

I was reminded a little of Aardvark, the social network search engine that Google purchased this past February, when I read this patent. Maybe elements of this rating system might be incorporated into the social network that Google is said to be launching sometime next year.

Share

73 thoughts on “The Evolution of Google Ratings (Experts Delegating Authority)?”

  1. I agree with Bill that this could be the future for Google. They are in the business of satisfying their searches and advistisers by providing high quality content to make it a win for everyone. I think people who produce content that passeds the spiders of the search engines and not their human visitors will loose out.

  2. I hope that google sticks with algorithmic calculation since this is completly based on objectiv science. Imagine the power of this “authorities” and the damage the could do when pribed by companies. Especially the pharmacy industry is a good example for this concern.

  3. I agree with Pascal, if Google were to switch to a Rating system it seems like it would open the door to all sorts of unethical behavior.

  4. Hi Bill,

    It’s possible that this is something that we will see come out from Google. I can see a service like this existing sider-by-side with Google’s present reviews that people might leave about their local auto mechanic or restaurants. Instead of competing with those reviews, these expert or social reviews have a somewhat different focus.

  5. Hi Pascal and Jason,

    I’m not sure that this kind of review and evalutation system would play a role in how Google might rank pages in its web search. I looked carefully through the patent to see if there was even some kind of insinuation of that possibility, but I didn’t see anything pointing to it.

    I definitely agree with you about the potential for bias and harm.

  6. I can definitely see a system like this emerging from Google very soon. Google is always changing and trying to improve how people find the information they’re looking for.

    I think it would be a great system, and as soon as becomes a hit other search engines like Bing will jump on board. It will definitely be interesting to see how they would implement this system though.

    If they could pull it off I think I would love it. It would definitely help reveal to the public who are the scammers and who are the ones that really know what they’re talking about.

  7. What an interesting idea. I wonder if this would prompt people to actually try and discover the truth behind things. It seems so many out there are just wandering along and believing everything that is written these days. Or would it only be used for very serious cases like the one you described?

  8. This is helpful in the sense that it will differentiate products or information from what’s factual and from what’s just hypothesis or something. And since they;re experts, their opinion will be respected as well as upheld highly and like all the others here who commented might really make way for unethical behavior. Maybe Google should look into it and pan ways on how not to allow it to happen.

  9. This seems like a pretty big burden for the reviewers. How would they be compensated? At some point, this could get huge, and I imagine there would be quite a backlog of articles for top professionals to review. It’s a great idea, but it seems hard to implement.

  10. I think this is highly likely and may already be in train. Google and microsoft have already confirmed that facebook + twitter have some influence on search results.
    http://www.seomoz.org/blog/google-bing-confirm-twitter-facebook-influence-seo
    I dont think its a huge jump of logic to think that they will include a peer review based signal in their algorithm in the near future. It will probably only have a small influence on ranks at the start, but over time and as machine learning improves this may grow. And who knows maybe they might use it in the quality score for ppc as well. A whole new SEO frontier is opening up….

  11. This is a very interesting article as usual Bill. it would be interesting to see how Google might use this to their advantage. The web is getting more and more reputation and review based, so this might compliment other sucvh tools like Wikipedia. As an SEO company based in the UK, with lots of reputation and added value for our clients I can see that this would benefit the web site owner and enhance their reputation. Good work Bill.

  12. I have a hard time believing Facebook and Twitter have any impact on serps. Facebook redirects any outbound links. Twitter nofollows links, most people use a shortner, and twitter is filled with possibly more worthless spam than any site in the world.

    Its also hard for me to imagine a voting system, because any automated voting system can be manipulated. Everything seems to cycle in and cycle out. look at digg.com. Cool in the beginning…now a worthless circle controlled ad server.

  13. Adam – Facebook and Twitter do not have any impact on SERPS. Facebook has actually been banned by Matt Cutts and his team at Google, so I am guessing that this will not change in the near future!

  14. I’ve seen numerous startup companies pop up like mushrooms after a rainy day, all trying to leave a mark on the ranking niche in one way or another. I guess Google noticed that as well :)

  15. This could be very dangerous for how information is conveyed. The biggest issue with the way we gather information now is that we are turning into a hive mind. Most people never really search for info any more.They just go to someone else. I remember when I was in college and many of the PHD professors (so called experts) were driven by very human forces. Sure they new the data and information but PHD’s are some of the most prideful, stubborn, people in the world (not all, but many). What happens if an incredible new theory is reviewed by an expert that has his own finding that would be challenged by the new paper? He may discredit the info. Very strict rules would need to be in play for this to work.

    On non academic articles, you would need real world insight as well. CEO, and upper management of companies are going to be a bit difficult to get to review massive amounts of information.

  16. Hi Kerry,

    It’s possible that Google may be building their way towards a system like this, but I predict that if they do, it’s going to cause a lot of headaches for the people working on it.

    The way that it’s described in the patent is that it wouldn’t impact rankings of pages in search results, but it would give people more to make an informed decision with, and that in itself can be a positive thing. I don’t think that it would be a good idea to replace the advice of a good doctor with a rating from an evalutation system like this, but it would be nice to see what someone from the American Diabetes Association might think of an article before you took it to your endocrinologist.

  17. Hi Susan,

    I would think that this system could be used for a fairly wide range of topics, from health issues to hobbies, and more.

    I’m not sure that we would be getting at the “truth” of matters, but rather the opinions of people who might have some expertise and experience in certain topics. Of course, like may of the comments following this post suggest, beyond that expertise, there may be other factors in the ratings that we might receive, even from people with expertise in a topic.

  18. Hi Andrew,

    In a way, I guess that you could see a system like this as Google’s attempt to provide peer review for publications on the Web, where there isn’t a peer review system in place. It’s not the same, and this evaluation system differs from that in a lot of ways, but Google does seem to be convinced that searches are very much interested in seeing reviews and ratings for a lot of things, from businesses to products, and in this case about documents found on the web (or even movies).

    I know that many professionals hesitate in giving out legal or medical or other advice in response to specific queries because there are often things that they don’t know about that may affect certain situations. I think that would be one major impediment in getting groups like the AMA or the ABA to be involved in a system like this review system.

  19. Hi Elizabeth,

    There doesn’t seem to be a compensation system set in place, or even one hinted at.

    I would guess that the impetus behind getting involved in a system like this would be that it could act as a channel to help an organization interact and communicate with the public in a positive manner, helping to advance awareness of those organizations and their missions. Many groups count amongst their business models or mission statements an educational aspect to what they do and how they operate. That might be the reason why a group like the American Medical Association might get involved.

  20. Hi Adam,

    What we seem to have heard over the past week or so is that social network activity and links may have some influence at the major commercial search engines. Maybe not in boosting rankings from things like PageRank, but they do seem to have some influece in what might be shown to searchers.

    Imagine a Digg where you can ask specific organizations their opinions on things that you’ve found on the Web. A political article – ask for rankings from both left and right wing think tanks. A medical article – ask from established medical organizations and alternative medical practitioners. The ratings would be from both members of those organizations, and people that they’ve hand selected to rate things. In many ways, that’s pretty different from the anonymous raters and commentors at Digg. The focus is upon finding known raters with an expertise on a subject rather than just anyone who signed up to vote on an article.

  21. Hi Chris,

    We don’t know if the links in places like Twitter or Facebook have any impact on other pages on the Web, but the content from places like twitter do showup themselves in search results at Google. How do those get ranked?

  22. Hi Dan,

    One of the things that Google can and does track to provide personalized and customized search results is the browsing activity of people using the Google toolbar and other Google services. I guess that gives them the chance to do things like notice how many people are actually using Facebook and Twitter and Groupon and other newer services.

  23. Hi des,

    The peer review analogy is something that I’ve thought about as well. There are some aspects about the peer review process itself that is limited, but the same can be said for the citation analysis approach that Google partially bases PageRank upon.

    It will be interesting to see if the power of a system like this grows upon how well something is ranked in search results, but I’m concerned about how it may bias search results as well.

  24. Hi David,

    The movie reviews aspect described in the patent reminds me of how I (and likely many other people) would look at DVD boxes to see how many thumbs up a movie might have received before renting or buying that DVD.

    The Diabetes paper I mentioned in my initial post is one that in many ways challenges the way that people manage diabetes, and hints at the possibility of an effective approach for managing the potential harm of high blood sugar that doesn’t make pharamceutical companies very much money. The author may be very well known and respected, and involved in organizations and research that are also very well respected, but it’s possible that “respected” rater might not give his paper high ratings because of some financial or political bias. In many ways, it’s exactly the situation that you describe.

    I think you’re right that there might not be enough incentive in certain areas for experts in those fields to share some of their knowledge and wisdom and experience as well.

  25. This is not good as far as I can tell. Best to stick with the science and have the Internet be a bit messy. Similar ‘authorities’ are government agencies like the FDA (google FDA, Monsanto and Taylor). Powerful moneyed interests will end up controlling these authorities and using them to peddle their own wares. There are endless examples of this type of corruption. Climate Science is another quick example.
    The framework is top-down: Finance, Corporations, Government, Me-and-You. IMO being on the bottom is not appealing.

  26. Hi Bill
    No that is true but what we do know is that there is no seo scoring factor for Twitter or Facebook, so they do not affect SERPs. I think that LinkedIn and Twitter are great for reputation management/enhancement and that Facebook may bring a community to your business, just not SERPs.

  27. Hi Mike,

    If organizations like the American Medical Association, the Nature Conservancy, the AARP, and so on were given a chance to be considered primary authorities for a rating system like this, it’s possible that they might abuse their authority, but it’s really in their best interest to act as if they are authorities. They would have the potential to enhance their reputations by being considered authorities, while harming that same reputation if they abuse that authority.

    Do you think Google would be showing too much faith in designating certain organizations as authorities?

    The patent itself doesn’t describe the process by which an organization would be determined to be a primary authority, or how much “authority” they might be given to delegate amoungst others, to make this more scalable.

  28. Hi Chris,

    Danny Sullivan, of Search Engine Land recently wrote an article on how social signals from sources like Twitter and Facebook might influence search results, and a good part of the article involved getting some actual answers from both Google and Bing about those signals and how they may impact Web search results. You can find it here:

    What Social Signals Do Google & Bing Really Count?

  29. Good theory, but very difficult to institute in practice. There is simply too much content on the web to approach SERPS this way. Even if this could be instituted, the results probably wouldn’t be much different than they are anyway. The system that you’ve described works very well for small niches (like SD,) but as a practical matter it would be next to impossible to assemble a staff to do this.

  30. Hi Kristinn,

    I’m guessing that you’re concerned that there might be bias in the ratings given by authorities.

    The patent doesn’t tell us that these ratings would be part of how pages would be ranked, and instead notes that the ratings would potentially be something seen as annotations that can be associated with certain documents.

    So, if you asked for ratings on an article on the Protection of Wetlands from someone from the EPA and the Army Corp of Engineers, as well as someone from Green Peace, Earth First, and the Nature Conservancy, chances are very good that you would get a very different batch of ratings from each organization, regardless of what the article actually says. The point behind this evaluation approach isn’t to change or influence search results, but rather to get expert opinions from organizations that might be considered authorities.

  31. Hi Bobby R.,

    A couple of the main assumptions behind the patent’s approach is that (1) there’s some value in getting opinions from primary authorities, or experts, and (2) providing those primary authorities with the ability to delegate certain amounts of their “authority” to people they trust makes this approach much more scalable to a wider system than if they could only vote on their own.

    Again, this patent has noting to do with search results, and wouldn’t influence the rankings of pages in search results. What it would do is give the raters a chance to add their opinions about certain documents to annotations of those documents on the web.

    Don’t know what you mean by “SD” as a niche on the Web.

    Why would it be impossible for a number of “primary authorities” to delegate their authority for ratings?

  32. It seems an interesting idea, but I’m trying to get my head around how improvements need to be made to have different mechanisms for different types of content. Research information being found valuable by peers in their field – that’s a vote for the web content itself. Whether a local tradesman did a good job for somebody who found them through the internet. Same idea, but authority in what they actually do rather than their website. Search engines are trying to be all things to all people, it’s going to creak. Then you factor in the manipulation that can happen very easily today. Bogus negative reviews by competitors for example. But interesting stuff, will visit from time to time to keep up with thoughts. Keeping ahead of the curve is becoming difficult for sure.

  33. Rarely do I comment on a blog post with where the owner of the blog dedicates so much time to comment response. Well Done, you have earned a well deserved twitter follower @OrlandoSEO

  34. Thank you, David,

    I really enjoy hearing the thoughts and opinions of others on things that I post here, and I think it helps to interact with people who do spend time and thought providing comments.

  35. Hi Mike,

    I had some of the same questions, and spent some time thinking about the reviews that Google presently offers. I don’t think that Google’s mechanisms for reviews for products or businesses are going to change in any way, or be involved in this Authority rating process. Google does seem to have some ideas on how to manage those types of reviews. See my post How Google may Manage Reputations for Reviewers and Raters.

    The expert rating approach does seem to be more geared towards documents found on the Web. I’m wondering if they should have even mentioned that this approach might be used for other content, such as providing ratings for movies. The patent doesn’t really describe how that might work, which is a little troublesome, especially since the document ratings can be explicitly tied to a specific URL, but movie ratings can’t.

  36. I think it’s potentially a sound idea, providing it’s not used as a factor in actual rankings, otherwise it could, as people say, open a can of worms! But it could work in tandam nicely, giving users a guide to what real exterts think on subjects.

  37. Thanks for your response. I always have one eye out for propaganda because it is so pervasive. However, there is no reason why these changes cannot be valuable sources of information. Annotations is a good way of looking at this, IMO. Sorry for the late response on account of my spam filter.
    Cheers

  38. Hi Dan and SteveF,

    It’s hard to say whether or not these “expert” ratings might ever play a role in the rankings of pages in organic search. The idea of ratings being provided as extra information, from valued sources adds a useful layer, but how much weight should a rating like one of these actually be given? If there are hundreds or thousands of ratings (is that even likely) from authority sites, is that better than a handful? Would it be more trustworthy? Should older ratings be considered, especially if the views about certain papers might change over time? Is there more or less of a possibility of bias and prejudice in ratings from experts than there are from anonymous reviewers?

  39. Hi Kristinn,

    Is it possible for an established organization to offer ratings without attempting to promote some specific point of view or cause? If everything that an “authority” rates looks like it’s according to a specific agenda to push forth some viewpoint, does that devalue their own ratings? I guess we should probably expect some of that from certain groups – like an environmental organization promoting green approaches, for instance.

  40. Hi,
    I hope I am not sounding too cynical. When I search in a topic of interest, I try to get a variety of opinions, pro and con, along with different approaches to the subject (what would Freud say about this brand of crackers). I look at the merits of the arguments and make my decision based on the quality of the argument (at least I like to think so).
    Having bias in natural. For me being able to read/find the dissenting view is important. So, maybe along with the authorities, they can have the anti-authorities!
    On a side note, people do not devote much attention to any particular topic and their impressions and opinions are made quickly, so the name of the game is getting the message out first while blocking out the others if you can. So, on any topic, the first read review will have the most impact, on down to 3 or 4 max. After that it is game over for winning hearts and minds. So, the effects will be of these changes will depend to a great extent on how they are visually presented.

  41. Hi Kristinn,

    I think that’s a good approach – getting opinions and viewpoints from a wide range of perspectives.

    I think authorities on a subject can be chosen from multiple perspectives. For example, the Environmental Protection Agency would probably present very different opinions on a topic than The Environmental Defense Fund, and yet both groups are very authoritive.

    We really don’t have much information from Google on how they might present these rankings, but I think you’re right that the way they are presented can make a difference.

  42. Pingback: SERPD
  43. Hi Bill
    All very interesting stuff! I actually stumbled upon your site, excuse the pun, and found the reference to diabetes intriguing! As a Diabetic who has a diabetes related website that features information and news such as those mentioned in your opening post…. including the article on Vit D! I would welcome a ranking system along these lines, although as some have said, its open to abuse. The thing that bugs me, and this I guess is personal, is the amount of “experts” out there who preach on diabetes but are not diabetic themselves! This is an obvious statement I know.. But when you live with this condition you clearly develop a sole deep understanding of its workings and therefore in your own way are in fact an expert yourself. Sure, you dont have the degree, but you have life experience which is often more viable… I feel. I suppose then it would be nice to open up this ranking system not just to academics but also real people with a validated option.

  44. Hi Ross,

    The article on Benfotiamine was pretty intriguing when I first ran across it, and I spent a lot of time looking for other articles to corraborate and validate its findings. Again, the patent described here mentions the use of these expert ratings not as a way to increase or decrease the rankings of pages in search results, but rather as a way that people who run across articles can get second opinions from authorities who might be experts on a subject.

    The patent does also mention that a rating system like this might open up to ratings from people in your social network – so if you’ve connected to a lot of other people who have diabetes, or run sites related to diabetes, you might use a system like this to ask for their ratings on an article like the one I’ve written about.

  45. I believe that the ranking systems already in place are sufficient without big brother determining who lives or dies. Readers make comments and rate the material in forums, blogs, etc.. so isn’t that enough?

    John

    SEO Expert and Marketing Manager

  46. Nice work Bill. I’ve been assuming for some years that Google has to tread this kind of path of adding social identifiers to ranking factors. There’s multiple types of information out there that needs different types of authority delegation, and recognising the experts in an arena helps. For example, I’ll place a high value on your views on SEO, but I don’t trust you as a guide to restaurants or other resources where I live – you aren’t local enough to hold a valuable opinion about the best indian restaurant.

    This patent weakly confirms, for me, that Google is considering the “social graph” of who-knows-whom, within fields of expertise. And that, with “earned impressions” (mentions from plausibly un-influenced third parties in social media), will help get away fron the intrinsically corruptible system of purely link-graph based ranking. Search engine results can’t ever be entirely clean; the best you can hope for is that it won’t be substantially subject to undue influence… But that’s a whole other story about regulatory oversight for a system that relies on human interpretation to create and select the algorithms used!

    Why only weak confirmation? Timing. What is intriguing is that it is dated 2004 – meaning that the corruption of the link graph may have already been considered (spamming being a strong problem at that point, IIRC), or that this is really part of the defensive armoury of patents to prevent another SE from using high quality data that might threaten Google. It’s been added to the portfolio at IPO time – is this just an investor protection mechanism, rather than an important search mechanism?

  47. Hi Bill,

    Thanks for the reply, I understand a little more ref the thought behind this now and to be honest, I would welcome a system such as this. Although I guess it just an added bonus rather than necessity… As a side not, I took Benfotiamine for a while as fat soluble form and found it excellent for nerve repair.. Had a bit of numbness on an ankle after an injury, sorted it right out! Bit pricey though. Shall keep an eye on your blog now I’ve found it, all very interesting discussion.

    Kind regards

    Ross

  48. Hi John,

    We’re not really completely sure what ranking signals are in place at Google, are we? We know about things like PageRank, and some other ranking signals that the search engines have told use that they use, but there are likely others that we can’t be sure of completely.

    It’s possible that there may be some bias in search rankings – actually we know that there is in some areas, such as web spam. We also know that a system like PageRank, and other algorithms based upon link analysis tend to favor sites that are more popular, and are linked to by more popular websites. Without that use of link analysis, we would probably have to look through a lot more pages in search results to find pages that may be relevant for our queries.

    Search engines do have some control over what we see, and we could probably make some great analogies to “Big Brother” when it comes to that control, but without them exercising some level of control, we would likely have much less relevant results at the tops of search results.

    People do have plenty of places to provide rankings and ratings and express their opinions about products and services and the quality and value of articles, but if we don’t have a way to find those, there’s a limit to how valuable they are. With a system like the one described in the patent filing, we actually know which organization is providing the ratings – there’s no anonyminity in the system, and there’s some level of accountability as well – these organizations are putting their reputations at risk by making ratings.

  49. Hi Jeremy,

    Thank you. Great points, and I think your article is something that anyone reading this post and wanting to dig deeper should read.

    A fairly intriguing book published in 2006 on link analysis, “Google’s PageRank and Beyond, The Science of Search Engine Rankings,” by Amy Langville and Carl Meyer, refers to the use of link analysis as an example of the use of the Wisdom of Crowds in ranking web pages. The crowd involved are web publishers write about and link to sites online. There are a mixed number of reasons why someone might link to a particular site or page, but largely we don’t know much about the motivations behind most of those links. The book does note that one of the problems with such a system is web spam, and the use of links only for purposes of increasing rankings, which is one of the biggest challenges that a search engine that relies upon link analysis for rankings faces.

    Your article does a great job of describing why, for some sites and queries, advertisements may provide more relevant results than organic listings – the people advertising are known, they pay a price to have advertisements shown, and therefore we understand better why the pages displayed in ads are shown where they are, because we know more about the factors influencing their inclusion, including an actual economic cost for their being listed by the advertisers.

    I don’t think that the primary purpose behind a patent like this one is more a message to investors in pre IPO days than it is an interest on the part of Google to find additional signals to improve the quality of search. While the patent was filed in 2004, chances are under the patent system that it wouldn’t be publised until at least 14 months later, and patents have never really been the greatest way to get information out to a wide audience. When it comes to search related patents, there really weren’t too many people back in 2004 who were writing too much about the patents that search engines were applying for.

    Google has been working to expand width and breadth of signals that it looks at to rank pages, including social signals for a number of years. Amongst their patents and whitepapers, we see a good number that look at user behavior in searching and browsing and providing annotations to consider rankings. These signals expand beyond looking at the activities of web publishers to web users to decide upon rankings.

    There are also patents filed from Google that attempt to look at the expertise of individuals, instead of just the “wisdom of crowds,” like the series of patents from Ramanathan Guha that look at labels and annotations in places like the context files associated with Google Custom Search Engines and possibily other systems of annotations, and include Google’s Trust Rank algorithm.

    This patent does describe how a system like this can also be applied to individuals within a social network in addition to public entities. From the patent (my emphasis added):

    A primary authority may be a public entity, such as the American Medical Association, or a private entity, such as an individual with a trusted Web presence, a peer of the user, or the user himself. Preferably, the primary authority designates contributing authorities that it believes hold opinions consistent with its own opinions. Likewise, contributing authorities preferably designate additional contributing authorities with similar views. The delegation of authority thus ensures that although the primary authority may not directly evaluate a portion of content, the rating determined for the content is reflective of the opinion of the primary authority. Viewed externally, then, the composite rating obtained from the evaluation system represents the value of the content as if directly evaluated by the primary authority.

    So, in addition to asking for known experts or authorities on a subject, this system anticipates that you may want to specifically ask a friend who you believe might have some expertise or authority on a subject – for instance, someone you know who might be from a certain area and tends to dine out a lot when you have questions about a guide to local restaurants that you’ve come across on the Web.

    There may be the possibility that ratings like this might someday be used to influence rankings of web pages, but the patent only notes at this point that they would be used as annotations for documents found on the Web. I would think that ideally Google would collect this kind of data, and then experiment with it to see if it does add value in rankings. The information itself as annotations add value to Google’s search results, even without a ranking influences – to a degree they would mitigate some of the problems of rankings based upon a “wisdom of the crowds” approach by providing some “expert voices” to temper those.

  50. Hi Ross,

    You’re welcome. At the time that this patent was written, it seems that Google anticipated the use of ratings like the ones described in this patent as a chance for searchers to get more input from people for something that they found on the Web from people who may have some expertise on a subject. As I noted in my comment above to Jeremy, we don’t know if Google would at some point use ratings like these as a ranking signal – I think they could possibly wait until they had a reasonable number of ratings before even experimenting with the possibility that such ratings would add value, and wouldn’t be abused by people trying to find another way to manipulate web page rankings.

    I’ve personally found some value in the use of Benfothiamine myself as well, and in the use of Methylcobalamine (a synthetic version of B12), but I would have loved to have been able to use a rating request system like the one described in this patent to ask for the thoughts of some experts in the medical field.

  51. This scheme reduces the room for “legitimate emergent newness” — cutting edge contributions sought by people who have already been the route with conventional expertise, found it lacking, and are looking for something else.

    The problem is that so much that is out there consists of half-baked ideas verging on superstition or, not much better, half-baked quasi-scientific ideas.

    The experts might be good for filtering out the latter, but not for filtering in the former, as steeped in and biased toward their own ideas.

    The greater virtue of the ‘Net is emergent newness. Pity if that were lost.

  52. I don’t know whether this system is the answer, but we definitely need something like peer-review on the web. Social media is providing this to some degree by highlighting the most influential people in an area of interest. It would really be good to have some way of evaluating the worth (or truth potential!) of pages published on the web. Cut down the noise and misinformation that’s out there somehow.

  53. This is big brother at work. Teams of wealthy, profit-oriented groups will control the flow of “acceptable” information. Beware my friends, beware of all so called “experts”.

  54. Review sites that offer the ability for customers to post about their experiences at a business are becoming more beneficial to users who may be faced with multiple options.

    The interaction between the company being reviewed, and the customer also carries weight because it indicates a particular business is interested in ensuring a quality experience.

    The manner in which businesses keep accurate information on the sites they’ve been listed on will help their SEO. A company that doesn’t take the time to keep this data current is likely to also provide the same type of accountability to the customer or visitor on their site.

    For small businesses, the best advice is to treat online visitors as you would someone who walked into your location. Make a good effort in building quality content and participate in comments if its on your blog.

  55. I can definitely see an issue with this, people are going to abuse their authority. Google I think may consider only a part of this, it’s just too open to abuse, and unfortunately the web is the worst place of that.

  56. Hi Lawrence,

    I’m not sure that I see this approach from Google as one that imposes limits on our discovery of the new, the different, and emerging technologies and philosophies.

    It may add a layer of annotations to results for pages in search results, and it may provide opportunities for established authorities to rate pages upon the request of searchers, but it doesn’t impose upon the web an organizational scheme in the same manner that many electronic databases are organized, such as something like the dewey decimal system. Someone requesting ratings can turn to a public authority who may be participating, or they may turn to a private authority – a peer, a friend, a connection in a social network. But, as it stands now in the patent, those ratings won’t influence the rankings of pages in search results.

  57. Hi Glenn,

    It’s actually not a surprise that a system like the one described in the patent might appear to be similar to peer review. A lot of the indexing approach that has been developed by Google in the past has roots in academic approaches like PageRank being roughly modeled after citation analysis for scholarly papers.

    One concern that I have is that popularity is often mistaken for quality – I think that’s a mistake.

  58. Hi webmind,

    I’m not convinced that the big brother analogy is a good fit here. Under this system, it’s just as valid to ask for ratings from your friends, or your peers, as it is from some public authority. When you ask for a rating, you choose whom you would like to rate a document, and the choices can include nonprofits, your neighbors, your friends, and others.

    Since the ratings aren’t anticipated to control the rankings of pages rated, it isn’t a mechanism for control over the flow of information.

  59. Hi Scott,

    Thanks, but I’m not sure that you read the post above, or skimmed through the patent that I linked to within it. The “ratings” described have nothing to do with reviews or ratings for businesses or products. Would love to hear your thoughts on the post itself, if you return.

  60. Hi Andrew,

    Chances are that some people will attempt to abuse their authority in a system like this, just as people have been atttempting to spam reviews for local search, and creating content and link spam in attempts to manipulate rankings for web pages. If a system like this is set into place, I would expect that there would be some kind of mechanism created to try to identify when there are problems. For example, if people post annotations to pages via Google’s Sidewiki, others who read those annotations have the ability to rate those annotations. It’s possible that Google might reduce or remove the authority of organizations and people who abuse a system like this.

  61. I think Bill is right. Google is taking some extreme measures, Lets hope the culprit ‘s only get hit by teh changes not everyone

  62. Hi Usman,

    I think this process has a chance of making the Web just a little better for everyone, by providing people with a way to ask for help from others. I’d love to see it happen.

  63. I also agree with Bill that a system like this will work well when implemented correctly and policed appropriately. Social media and the entire Web Content Management Software industry is doing some of its own policing, but has a long way to go toward these goals. Is this shift likely to occur over many years to come?

  64. Hi Michael,

    We may or may not see a system like this developing in the near future, but it’s interesting to think of the possible implications of such a system. What would it take to implement it? How much would organizations participating be trusted? How would it be policed? What kinds of steps would have to take place for Google to move forward and develop it?

  65. Couldn’t they also just tweak the pagerank passed to be somewhat more user-definable. I could to pass 40% of the pagerank to x link and split up the rest to all links. And adding it to be specifically targeted to keywords wouldn’t be that hard. The bigger problem, I would think, is getting people to spend their time doing that. So somehow meta tools would be needed to make this practical. I am not sure how they would work but I could imagine some such system working pretty well.

    There would obviously be all sorts of gaming risks that would have to be dealt with somehow.

  66. Hi John,

    The process in this patent filing really isn’t about pagerank or some kind of personalized rankings of pages, though I do agree with you that is likely something that Google could potentially do, and we see descriptions of something like that in whitepapers from Google going back to some of their earliest.

    The focus of this patent is more on creating the ability for people to get recommendations from both people within their social networks as well as people who might have some expertise on the topics being reviewed/recommended.

Comments are closed.