YouTube automatically recommends videos through a machine learning algorithm that analyses the characteristics of videos and the behaviour of its 1.5 billion users. Recommendations now drive 70% of overall “watch time” on YouTube, compared with 40% in early 2014, the company said. Now YouTube is gauging satisfaction by surveying nearly 10% of users about which videos they enjoy. The company releases neither recommendation nor satisfaction data about individual videos. Next year, YouTube is planning to have a similar initiative around science videos to surface “the established belief on the topic” on science videos, she said.

Source: Mint November 29, 2017 10:41 UTC