Audience Participation… System Implementation Level ---- Data Model Level ---- MachineRackData CenterInternet

Embed Size (px)

Citation preview

Audience Participation System Implementation Level ---- Data Model Level ---- MachineRackData CenterInternet Ambient Data Information Production Insights & Actions $1.10 $1,000 $1,000,000,000 $0.00 Source:$December $660M/TB August $100/TB Digital Shoebox Source Traditional Systems Data Warehouses / Marts Cubes Traditional Systems Data Warehouses / Marts Cubes Emergent Systems Deep data mining Machine Learning Near real-time prediction Emergent Systems Deep data mining Machine Learning Near real-time prediction Time Question Collect the data Build a logical model Build a physical model Load the data Tune Answer the question Question Worth asking again? Make it repeatable Bring it to production Validation Different Question Not interesting Source T1T1 T2T2 T3T3 T4T4 T5T5 Tree of transforms and filters Cleansing often happens in transformed domain E.g. Where I slept each night Can produce higher level information [DwellAtHome],[RouteToWork], [DwellAtWork] = Commute to work Using higher level information: Commute duration f(leavingTime) :18:26, :16:18, :21:18, :27:50, :24:37, :43:58, :26:48, None, :29:37, :53:34, :34:41, :00:25, :39:52, :44:54, :43:18, :28:49, :18:26, :16:18, :21:18, :27:50, :24:37, :43:58, :26:48, None, :29:37, :53:34, :34:41, :00:25, :39:52, :44:54, :43:18, :28:49, Dwell geolocation Outlook statistics + = How muchdo I send from home vs. at work? Reduce 12345