time series effects for tv recommendations
TRANSCRIPT
TIME SERIES EFFECTS FOR TV RECOMMENDATIONS
Diana Hu Verizon Labs
RecSysTV, September 2016
HELLO!
○ Currently @ Verizon Labs ○ Formerly @ Intel Labs ○ Large Scale Machine Learning & Computer Vision ○ Scala & Spark since 2014
You can find me at:
@sdianahu
DATA @ IPTV VERIZON
○ Millions of subscribers ○ Live TV playbacks ○ DVR recordings ○ DVR playbacks ○ Time sensitive content by TV Schedule ○ Hundreds of channels: Long tail
TV EXPERIENCE
SOME HISTORY
first second
http://www.brucesallan.com/2012/10/13/evolution-technology-television-back-day/ http://avdesigns.com/blog/design-your-own-at-home-sports-bar-with-a-multiscreen-tv-system/
CREATURES OF HABIT
CYCLICAL PATTERNS
○ Prime time ○ Weekend binge ○ Sunday night football ○ Basketball season ○ Effect of Sitcoms ○ Cartoons ○ News
TIME OF DAY VOLUME
Time Segment Day Segment
6 MONTH TIME SERIES
GENRE EFFECTS
MAKING SENSE OF TIME Besides pretty visualizations…
NAÏVE APPROACH
FEATURE ENGINEERING
○ Day parting ○ Weekday parting ○ Volatility spikes ○ Intraday volatility ○ Rank across time segments for: □ Channels □ Genres □ Programs
UNLOCKING PREDICTIONS
TIME AS A CONTEXT Enhancing recommendations
SOME APPROACHES
USE CASES
○ Predicting viewership ○ Personalization ○ Re-Ranking ○ Adapting to TV content “shelf life” ○ Seasonality ○ Trends
ACKNOWLEDGEMENTS
○ Luis M. Sanchez ○ Humberto Corona