The YouTube video recommendation system

James Davidson, Blake Livingston, Dasarathi Sampath, Benjamin Liebald, Junning Liu, Palash Nandy, Taylor Van Vleet, Ullas Gargi, Sujoy Gupta, Yu He, Mike Lambert
Proceedings of the fourth ACM conference on Recommender systems - RecSys '10

Add Remove

tag-as YouTube on 2/29/2020, 11:28:39 PM

excerpt Personalized Video Recommendations are one way to address this last use case, which we dub unarticulated want. 1 on 2/29/2020, 11:29:07 PM

excerpt Furthermore, videos on YouTube are mostly short form (under 10 minutes in length). User interactions are thus relatively short and noisy. 1 on 2/29/2020, 11:29:21 PM

excerpt The set of recommended videos videos is generated by using a user’s personal activity (watched, favorited, liked videos) as seeds and expanding the set of videos by traversing a co-visitation based graph of videos. 2 on 2/29/2020, 11:29:43 PM

excerpt During the generation of personalized video recommendations we consider a number of data sources. In general, there are two broad classes of data to consider: 1) content data, such as the raw video streams and video metadata such as title, description, etc, and 2) user activity data, which can further be divided into explicit and implicit categories. Explicit activities include rating a video, favoriting/liking a video, or subscribing to an uploader. Implicit activities are datum generated as a result of users watching and interacting with videos, e.g., user started to watch a video and user watched a large portion of the video (long watch). 2 on 2/29/2020, 11:30:07 PM

excerpt For example, recommendations account for about 60% of all video clicks from the home page. 296 on 2/29/2020, 11:31:36 PM