From the course: Building Recommender Systems with Machine Learning and AI
Unlock this course with a free trial
Join today to access over 23,200 courses taught by industry experts.
Filter bubbles, trust, and outliers - Python Tutorial
From the course: Building Recommender Systems with Machine Learning and AI
Filter bubbles, trust, and outliers
- The next real-world challenge is a tough one. It's called filter bubbles, and it refers to societal problems that arise when all you show people are things that appeal to their existing interests. Let's say you're building a recommender system for a bookstore. If someone buys a book about a topic associated with right-wing politics, your recommender system will probably pick up on that and start recommending more right-wing books to that person. If they respond to those recommendations, that person gets more and more immersed in right-wing ideology. The same happens for someone who bought a book about left-wing politics. The end result is that the recommender systems we develop to try and show people interesting things creates a more polarized society. This isn't something we ever anticipated in the early days of building recommender systems; but as the same techniques have been applied to social networks and advertising, it really has caused people to be exposed almost exclusively…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
-
-
-
-
-
-
-
(Locked)
The cold start problem (and solutions)6m 12s
-
(Locked)
Implement random exploration54s
-
(Locked)
Exercise solution: Random exploration2m 18s
-
(Locked)
Stoplists4m 48s
-
(Locked)
Implement a stoplist32s
-
(Locked)
Exercise solution: Implement a stoplist2m 22s
-
(Locked)
Filter bubbles, trust, and outliers5m 39s
-
(Locked)
Identify and eliminate outlier users44s
-
(Locked)
Exercise solution: Outlier removal4m
-
Fraud, the perils of clickstream, and international concerns4m 33s
-
(Locked)
Temporal effects and value-aware recommendations3m 30s
-
(Locked)
-
-
-