Rogers data days_2014_slides_opti
0 likes1,183 views
This document summarizes a critique of big data presented by Prof. Dr. Richard Rogers. The critique outlines 5 points: 1) A reputational shift to large infrastructures over other analysis modes, 2) A shift from interpretation to pattern recognition, 3) A shift in analytical commitments to visualizations, 4) A shift to relying on small data science teams, 5) A shift to relying on available signals rather than seeking unavailable ones. The document also discusses examples of big data hype declining and issues with tools like Google Flu Trends vastly overestimating influenza rates.
Rogers data days_2014_slides_opti
- 1. BIG DATA CRITIQUE Prof. Dr. Richard Rogers New Media & Digital Culture University of Amsterdam
- 2. Overview 1. Whither Big Data? (Wohin?) 2. Recent shift in Big Data from hype to disillusionment 3. Five Big Data Critiques 4. After Critique: Enriching Big Data Prof. Dr. Richard Rogers, University of Amsterdam
- 3. Google Flu Trends: Predictive algorithms in action Prof. Dr. Richard Rogers, University of Amsterdam
- 4. Prof. Dr. Richard Rogers, University of Amsterdam Google&Flu& Trends&vastly& overes2mates& rate&of&influenza& in&2012&and& 2013&in&the&USA& Is&Big&Data&measuring&societal& trends&or&media&effects?& Sources:&D.&Butler,&Nature,&494,&2013;&D.& Lazer&et&al.,&Science,&343,&2014.&
- 5. Big Data – a recent dip in hype and first indication of disappointment “Big Data” July 2013 * *July 2014 Source:&Gartner,& Hype&Cycles,&2014&
- 6. BIG DATA CRITIQUE FIVE POINTS
- 7. Big Data Critique – 5 points 1. Reputational shift to large-scale infrastructures and their analysts (Big Data handlers) • Big Data and Big Analytics are preferred modes of analysis and are awarded funding and contracts over other modes. Prof. Dr. Richard Rogers, University of Amsterdam
- 8. Big Data Critique – 5 points 2. Paradigm shift from interpretation (hermeneutics) to pattern-recognition, or from ‘close reading’ to ‘distant reading’ • Big Data focuses on overall trends and anomalies - the obvious and the obscure. Prof. Dr. Richard Rogers, University of Amsterdam
- 9. Big Data Critique – 5 points Twitter co-hashtag analysis of Dutch federal civil service use, 2012-2013. Visualization by Gephi (Borra and Rogers, 2013)
- 10. Big Data Critique – 5 points 3. Analytical shift in commitments and literacies to who can ‘read’ and grasp the visualisations • Big Data readings can prompt apophenia: seeing patterns where there are none. (boyd and Crawford, Info Comm Soc, 15, 2012.) Prof. Dr. Richard Rogers, University of Amsterdam
- 11. Big Data Critique – 5 points 4. Organisational shift to relying on one small team of data scientists (with special data access privileges). (See Tiropanis, Hall et al. Big Dat, 2, 2014) • “Thick data” specialists are no longer providing the insights. Prof. Dr. Richard Rogers, University of Amsterdam
- 12. Big Data Critique – 5 points 5. Shift to relying on (available) signals rather than seeking out (unavailable) ones. • Relying on more and more (available) signals does mean that there is good societal ‘coverage’. There are data-rich and data-poor (Crawford, Harvard Bus Rev, 2013) Prof. Dr. Richard Rogers, University of Amsterdam
- 13. Big Data – a recent dip in hype and first indication of disappointment Source:&Gartner,& Hype&Cycles,&2014& July 2013 * *July 2014 July 2015? * “Big Data” Next steps: Incorporating Broad Data, All Data, Thick Data and Small Data into Big Data
- 14. THANK YOU Prof. Dr. Richard Rogers Author of Digital Methods (MIT Press, 2013)