Search Engines and Serendipity

Search Engines
Image by Stuart Miles at FreeDigitalPhotos.net

Search engines are having a significant impact on the information that we consume. Indeed even articles like this one are influenced to some extent by how they may be retrieved by search engines — perhaps by using the ‘Search Engine’ term instead of simply ‘Search’ in the title. But let’s forget about the theory for a moment and concentrate instead on the practical effects and what they may mean for our ability to learn in the longer term.

Positive Benefits of Search Engines

Two elderly gentleman were browsing on iPads over their breakfast comparing  ailments and medications and browsing for information on related topics. There was nothing remarkable about their conversation but their use of the iPads highlights  how much the ways we obtain and share information has changed in recent times. The iPad in particular has delivered a user experience so intuitive that the hardware is no longer a barrier between user and content: it has taken away the fear.

Downside of Search Engines : How algorithms limit the information we retrieve

Without question, developments in technology have empowered end users but it is important to remember that behind the scenes algorithms — sometimes assisted by humans — are making decisions about the information that we see in our search results based on what they have learned about us from our previous Internet behaviour. This matters a lot when you consider that the gentlemen browsing in my local coffee shop are probably seeing search results biased by commercial interests and that they may be making decisions based on that information, perhaps even going so far as to self-prescribe and purchase medications online.

Personalisation

More worrying still is that every search and every click potentially feeds into a personalisation process that means the results presented come from an increasingly narrower set of potential results (which is one of the reasons why the EU Cookie law is important). Some months ago, I clicked on an ad for a dress. Since then, the ads I see most often as I browse the web are for dresses from the same supplier. You might think that doesn’t matter but what if I had clicked on  a political link? Would I  increasingly see results that reflect the particular bias of the original clicked link? Would I only see results from points of view that are similar to my own and what would that mean for my worldview? And what about those two gentlemen in the coffee shop? I suspect they may be seeing a lot of online ads for medications.

Loss of serendipity?

Attempts to deliver ‘quality’ and ‘relevant’ results  in response to often poorly defined search queries take into account factors such as the searcher’s location, previous search history and many more individual user behaviours so that the highest ranking results come from a much narrower base than most searchers realise. And, while some algorithms try to build in an element of serendipity or ‘chance discovery’, I can’t help worrying that the delivery of  ’personalised’ search results — which include content that has been manipulated to achieve a high search engine ranking —  ultimately diminishes those results and limits the potential for the searcher to learn.

Paid Content

Amazon, Netflix and iTunes have demonstrated that users will pay for access to content that meets their needs. Perhaps the time is coming when users will also contemplate paying for search — if only in those areas where quality and trustworthy results are most important.

For now, perhaps the best advice is to keep in mind that the top results returned on your search may well have been manipulated into that top position for commercial or other reasons. Try to remain curious when searching by clicking not just on the stories you agree with but also — at least now and then — on those that challenge your point of view thereby creating your own serendipity and maximising your potential to learn.