Your task is to analyze an existing recommender system that you find interesting. You should:
Perform a Scenario Design analysis as described below. Consider whether it makes sense for your selected recommender system to perform scenario design twice, once for the organization (e.g. Amazon.com) and once for the organization’s customers. Attempt to reverse engineer what you can about the site, from the site interface and any available information that you can find on the Internet or elsewhere. Include specific recommendations about how to improve the site’s recommendation capabilities going forward. Create your report using an R Markdown file, and create a discussion thread with a link to the GitHub repo where your Markdown file notebook resides. You are not expected to need to write code for this discussion assignment.
An existing recommender system that i find interesting is the one used by Netflix. According to an article by Busines Insider UK, “Netflix’s most important metric now is what you’re most likely to watch and stick with for awhile”.(http://uk.businessinsider.com/how-netflix-recommendations-work-2016-9). Hence, Netflix bases its recommendations on the frequency and length of time spent by a customer watching certain types of movies or tv shows.
The target users of this site are tv lovers, persons who love to spend some of their time watching a movie or tv show. The key goal of the target user is to be able to view a movie or show which appeals to their interest. The site strives to help users achieve their goal throuagh the current recommendation algorithms used.
If i am to judge the effectiveness of the recommeder system based on the experiences of users in my household, i would say that the system is very effective in perfectly recommending movies or shows that may interest the viewer. However, the limitation lies in the number actual movies and tv shows in their database that the recommender system has to reference. That number is insufficient. So in our opinion the amount of recommended movies and shows is insufficient. Some recommendations are also repeated across rows because they fall into multiple categories. So while the recommender system is effective based on its algorithm, in that it works well in being able to recommend movies and shows based on what a person regulary watches and the time spent on watching certain movies and shows, the system fails to present sufficient variety.
In order to improve the recommender system, i would suggest the following: 1.Allow for more variety, inclusive of allowing the user to add a movie or show name to the recommended list so that it can be made available.
- Avoid repeating recommended titles across rows. This can be annoying.
- Train the system to recognise recommendations that were previewed by the user and those that were ignored. The system can then ask the user to rate their interest in the titles they previewed. Recommendation can then presented based on the analysis of these metrics. For example, those previewed and rated high can be displayed first followed low rated, followed by new titles then followed by old titles that were ignored. This can demonstrate that the system is focussed on helping the target user meet their goals.