Google News Recommended For You
I was perusing Google News for inspiration for something to write about, when I came across the ‘For You’ section, where news articles are recommended based on your interests. I was looking through the articles recommended for me specifically, and could distinctly see the persona Google has taken me to be. My articles spanned psychology, job recruitment preparation, Photoshop, and movies. These were all things I tend to click on. I was intrigued by how Google put this list together so I looked further into it, finding this article on research into Google News recommendations: https://www.researchgate.net/publication/221607614_Personalized_news_recommendation_based_on_click_behavior
The article explained that for users that are logged into their Google account and have web history enabled, the recommendation system is able to build profiles of users’ news interested based off of their past click behavior. It uses a hybrid method of combining a content-based recommendation mechanism which uses learned user profiles with an existing collaborative filtering mechanism to generate personalized news recommendations.
I found this feature quite interesting and related it to network-based matching markets. The recommendation system matches users to articles based on their preferences and history, and creates a network of people with similar interests. It draws connections between articles and users, and gives insight to where the individual is within the network.
Sources:
https://news.google.com/?hl=en-US&gl=US&ceid=US:en
https://www.researchgate.net/publication/221607614_Personalized_news_recommendation_based_on_click_behavior