160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION The Power of Big Data Analytics Jaideep Srivastava Research Director QCRI Professor University of Minnesota [email protected] September 28th, 2016 Amsterdam, Netherlands 2 Talk Outline • Big data analytics • Digital response to humanitarian crises: a Big Data Analytics case study • Automatic analysis of food imagery • The power of big data analytics • Big data in food certification
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
ThePowerofBigDataAnalyticsJaideep SrivastavaResearchDirectorQCRI
September28th,2016Amsterdam,Netherlands
2
Talk Outline• Big data analytics
• Digital response to humanitarian crises: a Big Data Analytics case study
• Automatic analysis of food imagery
• The power of big data analytics
• Big data in food certification
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
3
Big Data Analytics - Definition
Big data analytics is the process of examining large data sets to uncover hidden patterns, unknown correlations, market trends, customer preferences and other useful business information.
4
TheKeyTakeaway
4
System
Environment
Little interactionwith environment
Lots of interactionwith environment
(Relatively)Static
Dynamic
Well solvedproblem
Well-formulatedproblem; needswider tolerance
(Mostly)don’t care
Most difficult case;becoming increasinglyimportant
• BigDatahelpsincapturingtheenvironment
• Analyticshelpsinmakingsenseofit
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
5
SocietalChallenge:Howtoleveragesocialmediainformationtoimprovehumanitarianassistanceduringcrises?
6
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
7
8
Project Collaboration with UN OCHA and the Standby Task Force.
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
9
10
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
11
12
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
13
Timeline2010 QCRI Created2012 Crisis Computing program starts2013 ● AIDR Development starts● MicroMappers Development starts● Uaviators Created● Typhoon Haiyan Response with UN
OCHA2014 ● MicroMappers - Coconuts for
Computer Vision● UNICEF AIDR SMS - Zambia● Typhoon Hagupit Response
2014 ● WoRDE prototype with
Translators without Borders● Cyclone Pam (Vanuatu Response)● Nepal Earthquake Response
2016 ● QSTP accelerator to explore
commercial potential● QSTP PoC program● UNESCO partnership● UNDP partnership talks
14
WealthofInformationisAvailableonSocialMedia(Instagram)
Username
LocationPopularityTimestamp
Hashtags
Comments
Description
Social network
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
15
UnderstandingFoodImages&Scenes
FoodScene FoodPlate
16
HowisInstagramaFoodStoreFront?
HomemadefooddeliveryservicesarepopularonInstagram inQatar.
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
17
The Power of Big Data Analytics
Hybrid collaborative intelligenceHuman computationMachine learningMulti-relational graphsTemporal evolutionary networksScalable data processing
Crisis computing
BehavioralHealth
Data sharing withcommunity
News
Urbaninformatics
For the funOf it
18
BigDataAnalyticsforFoodCertification• https://demo.greenfence.com
18
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
19
Key Directions in Food Certification Analytics• From post-facto analysis to real-time processing
• From post-mortem analysis to predictive analysis
• Big data è multi-dimensional, comprehensive view
• Account for external factors via social media data, news, etc.
• Automatic identification of unusual events (anomalies), emerging trends, etc.
20
TheKeyTakeaway
20
System
Environment
Little interactionwith environment
Lots of interactionwith environment
(Relatively)Static
Dynamic
Well solvedproblem
Well-formulatedproblem; needswider tolerance
(Mostly)don’t care
Most difficult case;becoming increasinglyImportante.g. GreenFence’sOpen Economy Platform
• BigDatahelpsincapturingtheenvironment
• Analyticshelpsinmakingsenseofit
160928 – SUMMIT 2016 DAY 2 – PLENARY SESSION USING BIG DATA IN CERTIFICATION
21
ThankYou!
21
Related Documents
