Importance of Recommendation
Since the beginning of humanity, we have stored data in various ways; Writing them down being the most common prior to the internet. Nowadays information is readily available with the press of a button. Anything you can imagine is digitized . The expansion of digital information creates a potential dilemma in our daily lives. The curse of choice hinder our ability to actually decide on something, because we fear there is a better choice else where. This problem has increased the demand for recommendation to levitate the millions of choices to pick from.
Billion dollar companies have been built off of recommendation systems. One of the most well known company that incorporates a recommendation engine is Netflix. We’ve all been there, scrolling through numerous movie titles, having no clue what to watch. Only to give up after the 10th page and searching up The Office and re watching it for the 5th time(Yes, I know The Office is no longer on Netflix). Often time more option lead to indecisiveness.
Recommendation systems aren’t just to help you decide what movies you should watch. They are everywhere, from the exact isle your favorite grocery store places its slice bread and peanut butter to the suggested books on Amazon
Amazon the hundred billion dollar E-commerce stop has products by the millions. They use a form of recommendation called Collaborative Filtering where they would suggest product based on previous customers purchases, hoping to up sell you. The basic idea being since you purchase bread and jelly another user also purchase bread, jelly and peanut. It would be a safe to assume you might want to buy peanut butter also.
Another Big company, Spotify, releases a new playlist every week for each individual user consisting of song you might like. Spotify’s recommendation system could be a good example of anther type of recommendation system called Content-Based Filtering where the recommendation are based on the content rather than other users. Music taste is very uniquely different from person to person. Such that person A and person B could have completely different taste in music. Content-Based filtering leverages the “content” of the song such as genre,rhythm, artist as a basis of recommendation. The goal being if you enjoy the music you would listen longer.
Recommendation Systems are great implementations to businesses and provide great benefit, as mention before, for the end user. From reducing the overload of movie titles to the viewer, the suggested purchases on amazon, and enjoyable music to listen to on Spotify providing an efficient and accurate recommendation will provide a better user experience.