In November of 2009, Google announced changes to the way people search when logged into their Google Accounts, in an Official Google Blog post titled SearchWiki: make search your own.
The changes allowed searchers to move search results up in rankings that they see when searching for specific queries, suggest new pages for searches for queries, leave notes on pages that show up in search results, remove results from search result listings, and see notes on search results that others have left.
If you regularly log into Google to search or access Google’s Gmail and you stay logged in to your Google Account while you search and select pages and browse other pages on the Web, the chances are that you’ll see search results personalized for you by Google based upon those pages that you’ve selected from search results, pages that you’ve browsed, and pages that you’ve bookmarked.
Using Searchwiki means that you can personalize your search results even more for specific queries, making it easier for you to refind pages that you’ve found before or include pages in a specific search result that you want to return to regularly. You can also remove pages that you don’t want to see in those search results.
If you leave notes on specific pages, others may be able to see those and take advantage of them as well. According to one Google Help page which explains the features of Searchwiki, other searchers shouldn’t be able to see your page that summarizes all of the rankings, deletions, and notes that you’ve made.
A patent application from Google was published last week which appears to describe one aspect of Google’s SearchWiki, the ability to add additional search results to the results that you see in response to a particular query, though it adds a collaborative aspect that isn’t presently part of Google’s SearchWiki (and may never be), that allows you to let others see the pages you add:
US Patent Application 20090094224
Published April 9, 2009
Collaborative Search Results
Invented by Douglas J. Ricket, Evan H. Parker, Marcin K. Wichary, and Kenson W. Yee
Assigned to Google
Filed October 5, 2007
Abstract
Methods, systems, and apparatus, including computer program products, provide alternative search results for a query.
In one aspect, a method includes transmitting a set of one or more search results for a query to a client device for presentation to a user, where each search result refers to a respective resource, receiving from the client device an alternative search result submitted by the user for the query, associating the alternative search result with the query, and storing in a repository the query and the alternative search result, where the alternative search result is transmitted with the set of one or more search results for a new search of the query.
Under this system, people could see notes or comments on search results that others have added or commented upon and vote for or against those comments.
This patent filing doesn’t fully describe all aspects of the SearchWiki system, but it may describe future possibilities under the SearchWiki system.
The announcement about Google’s SearchWiki on the Official Google Blog tells us that those changes won’t be seen by other searchers when we make changes in our search results.
Google might never add the ability to share results that other searchers added within SearchWiki. Still, the patent filing provides some examples of how they might be shared, such as allowing sharing of search results amongst people who work together within the same intranet system, where people who do share the same network might be able to see results that were added by others within that local network. Search results might also be visible to other groupings of searchers. Here’s one from many sections within the patent on how pages that one user adds to their search results might be presented to other searchers.”
[0043]In some implementations, the set of alternative search results for a query returned for a particular user may depend on one or more parameters, e.g., the language of the search system interface used by the user, the location of the user, or the membership of the user in a particular group.
In some implementations, a user submitting an alternative search result can specify the subpopulation of users for which the alternative search result will be available.
For example, if the user is providing an alternative search result that refers to a resource that is of interest to a particular department (e.g., engineering, marketing, or legal) or user group, the user can specify that the alternative search result will only be returned for users in the particular group, e.g., by selecting a checkbox included in the user interface 4050 for the particular group.
In other implementations, the subpopulation of users for which an alternative search result will be available can be automatically limited. For example, alternative search results submitted in German could be available only to users located in Germany or only to users who use a search system interface presented in German.
A user submitting a query can specify one or more parameters that affect the set of alternative search results returned for the query in some implementations. For example, if the user is submitting the query in French but is fluent in German, the user can specify or request that the user receives alternative search results for German users.
Some requests can be rejected by a system administrator or automatically denied by the search system, for example, if the user is not authorized to receive the requested alternative search results (e.g., alternative search results restricted to managers only).
It also tells us that pages added to search results for a particular query might also be added to results for synonyms for that query or to other queries that share one or more search terms with the original query.
Another aspect of the SearchWiki system is the ability for searchers to remove specific results from the search results that they see for specific queries. I described a patent filing from Google that describes how that might be done in February 2007, in the post Google on Letting Searchers Remove Pages from the Web.
Conclusion
I can see using SearchWiki to help me research specific topics, and collect helpful resources, leave notes on them, and remove results that I don’t find useful or interesting or helpful for queries related to those topics.
Google might explore data about how people use SearchWiki to learn more about searchers and more about the pages that show up for specific search queries. Again, Google told us in the SearchWiki announcement and helped a page that adding pages or moving them up or removing them won’t affect the search results that other searchers see.
Google might someday add a feature like SearchWiki to their enterprise search system in the future to allow searchers to share search results and notes about those amongst users of their enterprise search.
If you haven’t explored Google’s SearchWiki, you might find it worth a look. You may find that it helps make your personal search experience a better one.
I have wondered if Google will start using the SearchWiki data to determine in some part how the search rankings will play out? I know they collect and use a myriad of data to determine rankings and I think SearchWiki behavior could be very helpful to them in determining which sites are more highly respected.
It would be very interesting if they started actually sharing site comments. Basically each search result could have it’s own mini forum. Websites would have no choice but to focus on their users first as reputation would be paramount. The user moderated aspect could lead to spam and manipulation. So I’m sure Google will fully test data before fully incorporating such features.
Even I think that searchwiki data can play a role in natural SERP over a vast location or on a specific location .for eg:- people from one city saving same page using search wiki can give a signal that this page is useful for a specific city
I no longer see the searchwiki for any of my searches on Google. Are they planning on bringing it back?
Hi Agent SEO,
I’ve wondered the same thing myself. Google does collect a great deal of user behavior information about people’s searching and browsing activities for personalized search and their webhistory option. They also collect user behavior information about queries, clickthroughs, bookmarks, alerts in query user log files about searches, and track things like clickthroughs and other behaviors. The searchwiki information adds a whole different dimension.
Anything that Google can do to keep its search engine results relevant, whether it be computer-based, people-based or a combination of the two, is good for anyone who uses search engines.
I am intrigued by this notion of combining social with computational as an inforamtion finding methodology. I’ve long thought that thought processing bipeds are the best search engines because they can better understand the nuances of language. That said, I’m will Bill. For specific topics, a search wiki would be useful for discovering information albeit redundant of the social bookmarking sites that services this information need very well. What is the benefit to me to move search results around when new items are coming on line and the algorithms are changing and my personal adjustments may become like sandcastles after the tide has come in. Why doesn’t Google, or someone, create a game changing interface much like they did in 1997? Surely the plain box on the white page is not sacrosanct.
I can’t remember if I blogged on this, but if I didn’t I definitely considered it. Anyway, I had the same thought Agent SEO had, but the more I thought about it, the less likely it seemed to me. Adding this info into the algo opens up some serious concerns about trying to police/filter information. If click fraud was opening a can of worms, this would be like leaving a nuclear bomb on a playground. This would inevitably be terribly misused and spammed and I don’t know how they’d try combating it on such a mass scale.
I don’t see the point in personalizing search results, surely that defeats the object of the exercise. I’m not too familiar with this searchwiki concept though… can anyone enlighten me?
Hi People Finder,
Generally, I agree with you. But what about when someone searches for “java,” and it’s impossible to tell if they want the programming language, the island, or the coffee? Also, when a search engine returns thousands or millions of results, many of those beyond the ten or so that appear on the first page of results are likely relevant for the search. I think the results that people see can sometimes be overwhelming when they are good, and disappointing when the results are relevant but not what searchers are actually looking for.
Hi Marianne,
I do like that Google shows us some alternatives on their Google Experimental page. I’d like to see even more there.
Hi Adam Henige,
I imagine that spam and misuse is something that Google has considered in the development of Searchwiki, which is why when you remove pages or move a result up, you are the only one who is affected. There are some references to the possibility of people attempting to spam and manipulate results in the patent filing as well. Making Searchwiki only available to signed-in users is likely also a way of limiting potential harm. I suspect that there are people at Google who are studying at what users are doing with searchwiki, and exploring how it might potentially be abused by users.
Hi Adam,
The best places probably to start learning about the features of Searchwiki are Google’s help page, and their blog announcement, which has a video introducing Searchwiki.
Why allow people to personalize their own search results? One of the main reasons would be to allow people to easily refind pages that they thing are relevant for a search that they might return to a few times, or even more regularly.
I think that customizing Goolgle Search Result is only one the obvious steps leading to the real meaning of the Web 2.0. Web 2.0 will be the next Internet where everything will be perfectly tailored to your need because…you shaped it by yourself!
Hi William, I understand that it allows people to re-find pages, but surely that’s what bookmarks, favourites and cookies are for. A search engine should find the best pages for you based on what really deserves to be there – not based on what pages the user has viewed before. If I search for ‘country music’ on Google and select a few results as preferred results, then I do the same in 4 months time – I would surely be missing out on newer and potentially more relevant results?
Thanks for the link about the search wikis – I’ll check it out!
I don’t like SearchWiki much, simply because I’m always presented webpages of websites I would have tagged for other search results not related too. I prefer to see always fresh content. If I need to check my bookmarks, I can do that on my own … thanks 😛
Hi Web Talk,
Is the future of the Web where you shape what you see? It’s an interesting question that you raise. I think one of the strengths of the Web is the ability to be introduced to new concepts, new ideas that might broaden your experience of the world. The Web provides us with the ability to discover new things. Would we lose some of that discoverability?
Hi SEO Blog Malta,
I haven’t used searchwiki much, not because I like it or dislike it, but rather because I’d rather see nonpersonalized search results.
I would guess that the part of Google’s algorithm that focuses upon providing fresh results would still be in place under searchwiki.
I guess one of the reasons to put something like searchwiki out into the world is to see if people actually use it, and learn if it’s something that people take advantage of, and how they do. I’d rather bookmark things I find interesting, too.
Hi Adam,
One of the questions that I have about searchwiki is “where’s the wiki?” If the idea of a wiki is that people collaborate together to build a repository of information, why don’t people see the actions of others in moving pages up, or removing pages? What happened to the collaborative aspects that are described in the patent filing? Is this something that Google might roll out in the future, or are they wary of doing so?
If you select a few results as preferred results in search wiki, the new results will still appear, even if they might be below your preferred results. We’re both concerned about missing out on potentially new and interesting results that might be lost to us if we vote things up and never look past our preferred results.
One difficulty that search engines face is what is it that actually deserves to be at the top of the search results. Is it possible that the most relevant results for me are different than the most relevant results for you? If I’m a java programmer, visiting the island of Java, looking for a cup of coffee, and I do a search for “java,” should I see results for the programming language, the island, or the coffee? All three types of “java” deserve to be at the top of the search results.
Hi Jens Peter,
Thanks.
We expect a lot from search engines on the basis of typing a handful or less words into a search box. When we don’t know very much about the topic that we are trying to learn more about, and we don’t know much about it to begin with, it can be really difficult to decide which words to even begin using to search for.
I’m not sure that a better understanding of language and how people use speech will help there. I think that’s part of the reason why we are seeing more query suggestions in search results, and why there have been a number of whitepapers that describe how a search engine might attempt to provide a diverse set of search results to searcherss. In my last comment, I mentioned a Java programmer, visiting the island of Java, looking for information about the Java variety of coffee. If he or she types “java” into Google, the search engine might make certain that amongst the top ten search results they include something relevant for the programming language, something relevant for the island, and something on point for the coffee.
Does something like searchwiki help that programmer? If he used it in the past to display the best resources for the programming language, then the coffee results might be buried somewhere below those, or he may have even removed them from the searchwiki results. I’m hoping that sometime in the future Google shares with us some information about how people have used searchwiki.
Hi,
interesting post…but as the comments shows thoughts processing search engines are a thing of the future. Until they fully understand the function of the speech tree they will most likely be inapt.
Jens Peter