Issue 39 December 2012

Being the ever so “benevolent” company, Google wants to help make things easier for you when you use their search engine.  So, in 2009, Google introduced personalized search, a concept in which your search results are customized for you based on various algorithms that Google uses to compose a profile of who you are.  In essence, gone are the days where everyone receives the same results in the same order when they search on common keywords.  Your results are now unique to you. 

 

Google systematically tracks, amongst other things, where you are located, what you search, what sites you select (that’s right – none of your searches are anonymous or forgotten) and will even track the order of your searches to determine if there is any relationship between two searches.  Google is, to put it crassly, a data whore – collecting, storing, analyzing and manipulating data on all of us that use their product (and let’s face it, who doesn’t use Google?).  Conceivably, personalization is the biggest change that has happened in search engines and has become a nearly invisible revolution in how we consume information. 

 

In some ways, this is quite convenient.  If you are looking for a great Chinese restaurant to visit, Google will bring to the top of your search results all the Chinese restaurants in your area.  No longer will you waste time wading through unrelated results; Google has filtered the results dependent on your location.  And that webpage you visit on a regular basis but haven’t bookmarked yet?  Each time you search for it, Google brings that site to the top of your search results - pretty convenient.

 

But what if you are searching for a balanced viewpoint on a certain topic?  That is, you want both positive and negative results to read and review.  Suddenly, with personalized view, Google now only shows you the results that mirror your viewpoint based on past search history.  Put another way, if you have conservative views, you will continue to receive search results that are conservative in nature; the liberal viewpoints get pushed down in the search results or never get shown at all.  We become encapsulated in a filter bubble.  Your search results are no longer balanced in perspective.  Author of The Filter Bubble: What the Internet is hiding from you, Eli Pariser notes: “Personalization creates an information determinism – what you clicked on in the past determines what you will see next; a sort of web history you are doomed to repeat.”

 

Search Engine Alternatives

So what’s a researcher to do?  In order to avoid the filter bubble and provide private search options to the public, a number of different search engines have entered the search scene.  These are a couple of examples of great alternatives to Google:

 

DuckDuckGo www.duckduckgo.com

 

  • DuckDuckGo (DDG) doesn’t log any personally identifiable information.
  • DDG claims that it does not collect or share personal information, including browser cookies.  Other features include the inability to store search history and lack of selling of personal information to third party companies.
  • DDG claims that all users get same results because this search engine has no way of filtering results.
  • http://donttrack.us/ and http://dontbubble.us/ provide explanations of how DDG differs from Google and other algorithm collecting search engines.

 

Ixquick https://ixquick.com/

 

  • Ixquick is a wonderful choice because it employs metasearch technology, which means that researchers may receive larger number of results.  Users have an option between an advanced search, a global search and power refinements.
  • Ixquick does not share personal information with any third party, including other search engines and the provider of its sponsored results.
  • According to Ixquick’s privacy policy, “your SEARCH is not recorded, your IP ADDRESS is not recorded, your IDENTITY is not recorded, no TRACKING COOKIES are placed on your browser, and our SSL ENCRYPTION ensures that your ISP or hackers can't eavesdrop”.
  • Unlike Startpage (below), Ixquick pulls results from a variety of sources instead of only Google – this can be a good or a bad thing, depending on how much one enjoys Google’s search results.
  • Ixquick and Startpage have essentially the same design. Ixquick includes the same privacy features Startpage does, including the Ixquick proxy links in the search results.

 

Startpage www.startpage.com

 

  • Startpage is a sister company of Ixquick and is known as an award-winning search engine with an industry-leading privacy policy.
  • When you use Startpage, the search engine will not record your IP address, what type of browser you are using (Internet Explorer, Safari, Firefox, etc.), your computer platform (Windows, Mac, Linux, etc.), and it will not record your search words or phrases.
  • The only information StartPage records is an aggregate total of how many searches are performed on its website each day (a measure of overall traffic) and those overall traffic numbers broken down by language.
  • When you search with Startpage, the Web results are generated by Google. This offers you the search results and search features you may have grown accustomed to.  Search queries are sent through a proxy, which lets users browse websites safely and anonymously, without passing on any private, personally identifiable information to the websites they view.
  • Startpage also includes a proxy feature — you can open a page in Ixquick’s proxy directly from the search results. This is slower than normal browsing, but websites won’t be able to see your IP address. The proxy also disables JavaScript to protect your privacy.
  • Ixquick and StartPage support approximately two million searches per day
     

How to minimize the filter bubble

It should be noted that Google is not the only company that uses personalization: the entire business model of Amazon and Netflix is based on the concept of personalization.  Bing and Yahoo search engines also use personalization.  So, what can we as researchers do to mitigate the effect of the filter bubble? 

 

Here are a few suggestions:

  • Burn your cookies (not the chocolate chip kind but the kind that get stored on your computer and track the identifying features of your computer).
  • Erase your web history often (both from your browser and from Google); much of the personalization is based on your past searches and selections.
  • Ensure you are in private browsing mode on your web browser (InPrivate browsing in Internet Explorer, Incognito mode in Google Chrome and Private Browsing in Modzilla FireFox).
  • Use alternative search engines like DuckDuckGo which claim they provide private search functionality.
  • Be aware!  Ignorance feeds the filter bubble. Why give all your personal data and control over to corporate entities?

 

Further Information for your reading pleasure

The following links provide further information on the concept of search engine personalization and the filter bubble:

Eli Pariser’s TED Talk on the Filter Bubble: http://www.thefilterbubble.com/ted-talk

Google’s 2009 video on personalization: http://youtu.be/EKuG2M6R4VM

Google’s Good to Know information on privacy: http://www.google.ca/goodtoknow/

 

Personalization is here to stay. In fact, within the next three to five years, the idea of a web site that isn’t customized to a particular user will seem quaint (as noted by Facebook COO Sheryl Sandberg).  The key to solid research is ensuring a balance of views and opinions so we don’t end up consuming only information junk food and ignoring the content vegetables.

 

 

Liz Rejman is a Development Researcher at London Health Sciences Foundation and Director of External Relations for APRA Canada. 

 

Izabela Piasecka-Latour, MLIS, is the Manager of Prospect Research at Ontario CNIB.

 

Share this with your networks!