Readings 7: Suggestions, Recommendations, and Algorithmic Culture

21 07 2011

Hitting it Off, Thanks to the Algorithms of Love explained the effectiveness and ineffectiveness of certain online dating sites.  Most of the sites are based off of matching personality traits, which is a huge factor in the success of long-term romantic relationships.  I think the romanticized idea that you have to run into your ‘soulmate’ on some chance occurance and then go through the long process of getting to know each other to decide if it’s a good match is going to become outdated soon.  The idea that love is based off of intangible things like fate might be overshadowed by statistics that demonstrate what usually determines a successful relationship.  There may be hesitation to accepting that, but I think it would be more logical to find a partner through what science tells us usually works than the traditional way of finding a significant other.  Also, it was disheartening to read that some sites didn’t allow homosexual matches to be made because the algorithms were designed from heterosexual data.  I think the same algorithms would work for any kind of relationship.

Artifacts from the Future: Online Dating in 2020 was kind of surreal; it’s not hard to imagine a world where our geneteic make-up could determine our most successful partners.  The idea would bring a completely new meaning to the classification of people.  I’ve also heard, though, that appearance is 20% genetic and 80% nutrition; if physical attraction plays a role in the success of a relationship, then this might not work as effectively as I immediately thought.

Roommates Who Click would have been a lifesaver a few years ago.  I’ve had some awful roommates in the past and if I knew there was software that could match me with someone ‘normal’ by my standards, it would have been a godsend.  When you’re potentially living with someone for 4 years, it’s best to have similar personalities and habits.

Search Takes a Social Turn brought up a lot of different sites that work as filters to accomodate your taste in various forms of entertainment.  They were based off of what was trending within your ‘social loop.’  S0 a site like TunerFish collects data from what your friends are watching and suggests it to you; a neat idea when you consider that you’ll probably later meet up with friends and discuss what you’ve all been watching.  It’s incredible that these sites suggest things you’ll like on such a specific level.  I personally like to explore different shows, music, movies, etc. to develop my own taste and maybe stumble upon something unfamiliar that I like, but the aforementioned filters are definitely a great way to follow media.

The Song Decoders at Pandora – I’ve been a huge fan of Pandora and the way it filters music has always been astonishing.  The article really delves into what creates the ‘magic’ of its inner workings.  I’m a music major, so I’m familiar with looking at music on many different levels, but the detail that Pandora goes into is astonoshing.  Countless pieces of data are being compared and contrasted to define the ‘feel’ of a song which truly amazes me.  It almost worries me that music will soon be so systematically created that the ‘blood sweat and tears’ behind making great music will disappear and ruin ‘authenticity.’  I realize though that it’s just another tool advancing the pace of creativity; I don’t want to sound like the guy who called record players “infernal machines.”

Recommender System entry on Wikipedia gives a nice clear-cut description of what a lot of the aforementioned systems are really doing.  This is a much more generalized picture of what is going on, but I think it’s good to look at how they work at a basic level.  I didn’t understand some of the jargon in the previous articles, but this helped clarify what the recommender systems were doing to create suggestions.

How to Have Culture in an Algorithmic Age is basically telling us to not totally trust the algorithms.  Though they are great and helpful, they can never factor in every aspect of the context in which a song, movie, passage, book, etc. can be graded.  And in the case of Amazon, the dynamic of the algorithm is hidden, so it is impossible to tell exactly how it is interpreting the data we send it.  In the example they mention with tracking Kindles, one person may highlight a sentence because it’s inspiring, while another may highlight it because of a grammar mistake.  I think the author is recognizing the validity of recommender systems and at the same time acknowledging that culture will persevere with these algorithms.




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: