Social Listening for Richer Data, Unstructured Analytics
summary
Shows concrete pipelines: pulls Foursquare check‑ins and maps Michelin star density to pick a Barcelona street, streams tweets with an Azure Logic App into Dropbox files, and runs tone extraction then trains models in Azure Machine Learning (Matchbox recommender, R text hashing) and on HDInsight/HBase/Hive. Cites a 376‑mother study with 71% accuracy, stock sentiment models at ~65% predictive lift, and a retention model that triages calls to raise effective saves toward ~90 per 100.
AI-written summary · about
chapters (19)
- 0:00Social listening intro
- 3:12Social data business value
- 7:49Restaurant case study Foursquare Power Map
- 12:48Azure Machine Learning for restaurant recommendations
- 15:58SQL vs Azure ML mining
- 20:39Storing social data
- 21:42Twitter data collection demo
- 27:31Competitive advantage insights
- 30:47Stock trends from social media
- 35:04Marketing vs social analytics
- 37:56Election Twitter sentiment
- 40:25Ad placement with social data
- 41:10Big Data tools for listening
- 42:34Social tone vs sales correlation
- 44:22Azure platform advantages
- 45:52Key takeaways implementation
- 48:12Customer retention prediction
- 51:47Tone analysis tools demo
- 54:18Resources next steps
related talks
01h 05mSocial Text, Sentiment, and Tone Analysis
41mDaniele Montesi & Emanuele Viglianisi
01h 02mData lakes beyond clusters
18mLessons from Analysing all of Europe's Data
01h 04mTake Control of the Data of You
59mPredicting the future as a service with Azure ML and R
47mAI vs. Karen: Real-Time Customer Sentiment Detection at Scale
34mPast The Hype: A Better Way To Think About Big Data