Recommendation Engines: Reinforcing What We Already Like

Recommendation Engines: Reinforcing What We Already Like

Recommendation Engines: Reinforcing What We Already Like 150 150 adminxs

What type of music do you like? Punk, soft-rock, rap, pop, country? After answering this question think about how you were introduced to that genre of music in the first place. I love soft-rock and hip hop because it was presented to me by my friends and family. However, in an era where the internet has intertwined itself with the cultural web of our lives, recommendation engines are starting to tell us what we will like next.
While SEO website design involves structuring websites in a way that makes it easy for search engines to find them and thus increase the profitability of a company, recommendation engines try to provide you with things similar to what you already like based on your original selection in hopes that you will buy, listen or view those additional products.

Pandora is a perfect example. Its “music genome project” is based on our initial music choice and similar songs that fit its criteria. This is known as “collaborative filtering.” Pandora takes a song such as P. Diddy’s “Homecoming” and identifies a multitude of attributes about that song and rates them; maybe it’s the fast beats, the rapping interlude or echoing sounds in the background. With all of these attributes categorized and ranked properly, Pandora will then choose more songs with similar characteristics. They even take it a step further by encouraging user interaction: you can click a thumbs up, thumbs down or “I’m tired of this song” button to control what you listen to next. (Granted, you’re only allowed about three thumbs down in a row that lets you skip songs.)
Several other websites feature recommendation engines that are focused on identifying similar items to what you’ve already selected. After you watch a video on youtube, it immediately pops up with an array of more videos that feature the same type of content. I originally chose a video of a “cute puppy,” and then a video entitled “Cute/Funny Dogs” pops up immediately after.

This is also a common trend that can be seen on clothing sites that have been designed with ecommerce web site development. Banana Republic showcases a “you might also like” section directly under the product you’re viewing that gives you a selection of similar clothing styles. I picked out a delicate blue sweater with pockets and it suggests to me skinny work pants, a “ruffle wrap” and elegant silk dress.

Lastly, Amazon, much like most other book sellers presents to you a deal of three books you could buy instead of one. After choosing Lewis Carol’s Alice in Wonderland, Amazon suggests that I purchase a bundle of books that are “Frequently Bought Together.”

With all of these recommendation engines intersecting our daily selections, the real question is are we really choosing what we like or what’s already there? Adhering to the advice of these engines can effectively put us in a rut of interests by always choosing things similar to what we originally liked. While friends and family can introduce new and exciting songs, clothes, books etc. to us, these “collaborative filters” can only tell us what we might like based on what they already know about us and therefore limit the creative expression of our daily lives. Your Orange County wen designers advice is: be adventurous every once in a while and try something that’s completely out of your comfort zone. Try categorizing that, recommendation engines!