Social Data and Television
summary
Describes how social networks link to TV by collecting public data via streaming APIs to build a dataset of billions of tweets and user links, processed with Python and grid computing to produce show-level features used for recommendations and viewership prediction. Emphasizes limits: need national-scale events or big brands for signal, geography and gender slicing drains sample size, and the surprising payoff is check-in data acting like individual-level set-top box information without industry cooperation.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:59How does the data come into you
- 1:53Do you choose TV shows based on popularity
- 3:15Can you narrow down by region
- 3:54Advertising events
- 4:48Processing the data
- 7:12Checkin Data
- 8:19Wrap Up
related talks
10mHigh Order Bit
03mVisualizing data with JavaScript at Twitter
10mBig Data and the Hypocrisy of Privacy
20mDesigning for Big Data
38mData and Business
37mLarge Scale Social Network Data Collection & Analysis...
21mBig Data and You: Four Ways to Use Data Every Day -
41mDemystifying Social Network Analysis (SNA)