SlideShare a Scribd company logo
1 of 9
Download to read offline
Surprise, You’ve Got Friends!
Problem Definition:
  Want to present the serendipitous connections?
  How do we define and mine surprise?                                     Then
Aspects of Surprise:
  Uncanny vs Unexpected
  Surprise can be measured as the change in prior
  to posterior probability, is time dependent.
Approaches/Main Idea
  Explicit communities, e.g. flickr group, are also connected by
  implicit groups, e.g. personomy measures, co-location via geo tags.
  Can find two people at this moment, this place, with the greatest
  Explicit Distance, and smallest Implicit Distance, this may surprise
  or abhor these people.
                                                                           Now
Weighting
 Consider geo similarity. Inverse weight geo location
 by radius of bounding box, or “fame” of location.
Test Case
  Use geo tagged flickr pictures to find two strangers
  at a gathering that have a visited the same unusual
  location in the past.
A unit of surprise --- a wow
                                                 may be defined for a single
                                                 model M as the amount of
                                                 surprise corresponding to a
                                                 two-fold variation between
                                                 P(M|D) and P(M), i.e., as log
                                                 P(M|D)/P(M)
                                                  (with log taken in base 2)





 •
 L. Itti & P. Baldi, Proc. IEEE-CVPR, 2005

 •
 L. Itti & P. Baldi, Proc. NIPS, 2006
First approximation



    Took the flickr group ISWC

         1.  Extract the tag distribution of the event based on the photos in
    the group(E)

    
 2.
For each user:
           Took the user vector and computer the TF-IDF using the group
              factor as the IDF
    
 3.
For each pair of users
    
      Compute the cosine similarity
Results




The two “Closest - furthest” people in this group
are from the same research institute in Ireland, DERI
Base Case




The base-case, using all the information, gives the
same result, however the fist pass analysis was
naive
Plenty of signals that we wanted to take into account:

Network of friends (two individuals were friends in flickr)

All tags from all users

Geo Coordinates

etc etc etc
Some Applications

Dating Service

For Academics
Some Broader Issues

Surprise is time dependent, current signals from some SNS are highly
bursty, e.g. batch upload in flickr

  Connecting this analysis to real world sensors would seem to offer a
  method for extracting better information on calculating priors.
  Connecting across data sources certainly will

A Pleasure Device “Surprise Engine” may become a tunable mess, much
like current search engines.

More Related Content

Similar to Mining Surprise

Grindinger Group Wise Similarity And Classification Of Aggregate Scanpaths
Grindinger Group Wise Similarity And Classification Of Aggregate ScanpathsGrindinger Group Wise Similarity And Classification Of Aggregate Scanpaths
Grindinger Group Wise Similarity And Classification Of Aggregate ScanpathsKalle
 
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...ijtsrd
 
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...Kan Ouivirach, Ph.D.
 
IRJET - Object Detection using Deep Learning with OpenCV and Python
IRJET - Object Detection using Deep Learning with OpenCV and PythonIRJET - Object Detection using Deep Learning with OpenCV and Python
IRJET - Object Detection using Deep Learning with OpenCV and PythonIRJET Journal
 
120_SEM_Special_Topics.ppt
120_SEM_Special_Topics.ppt120_SEM_Special_Topics.ppt
120_SEM_Special_Topics.pptzaki194502
 
Sgg crest-presentation-final
Sgg crest-presentation-finalSgg crest-presentation-final
Sgg crest-presentation-finalmarpierc
 
Deep Learning for X ray Image to Text Generation
Deep Learning for X ray Image to Text GenerationDeep Learning for X ray Image to Text Generation
Deep Learning for X ray Image to Text Generationijtsrd
 
IRJET- Deep Convolution Neural Networks for Galaxy Morphology Classification
IRJET- Deep Convolution Neural Networks for Galaxy Morphology ClassificationIRJET- Deep Convolution Neural Networks for Galaxy Morphology Classification
IRJET- Deep Convolution Neural Networks for Galaxy Morphology ClassificationIRJET Journal
 
Emc 2013 Big Data in Astronomy
Emc 2013 Big Data in AstronomyEmc 2013 Big Data in Astronomy
Emc 2013 Big Data in AstronomyFabio Porto
 
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...Kalle
 
Comaskey_William_Poster_SULI_FALL_2014
Comaskey_William_Poster_SULI_FALL_2014Comaskey_William_Poster_SULI_FALL_2014
Comaskey_William_Poster_SULI_FALL_2014William Comaskey
 
Shadow Detection and Removal Techniques A Perspective View
Shadow Detection and Removal Techniques A Perspective ViewShadow Detection and Removal Techniques A Perspective View
Shadow Detection and Removal Techniques A Perspective Viewijtsrd
 
What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? Robert Grossman
 
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million Galaxies
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million GalaxiesAstronomaly at Scale: Searching for Anomalies Amongst 4 Million Galaxies
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million GalaxiesSérgio Sacani
 
IRJET - Direct Me-Nevigation for Blind People
IRJET -  	  Direct Me-Nevigation for Blind PeopleIRJET -  	  Direct Me-Nevigation for Blind People
IRJET - Direct Me-Nevigation for Blind PeopleIRJET Journal
 
LLNL_poster_15.compressed
LLNL_poster_15.compressedLLNL_poster_15.compressed
LLNL_poster_15.compressedYoungwook Kwon
 
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...Xiaohan Zeng
 
Machine Learning ICS 273A
Machine Learning ICS 273AMachine Learning ICS 273A
Machine Learning ICS 273Abutest
 

Similar to Mining Surprise (20)

Grindinger Group Wise Similarity And Classification Of Aggregate Scanpaths
Grindinger Group Wise Similarity And Classification Of Aggregate ScanpathsGrindinger Group Wise Similarity And Classification Of Aggregate Scanpaths
Grindinger Group Wise Similarity And Classification Of Aggregate Scanpaths
 
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...
Shadow Detection and Removal using Tricolor Attenuation Model Based on Featur...
 
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...
Extracting the Object from the Shadows: Maximum Likelihood Object/Shadow Disc...
 
IRJET - Object Detection using Deep Learning with OpenCV and Python
IRJET - Object Detection using Deep Learning with OpenCV and PythonIRJET - Object Detection using Deep Learning with OpenCV and Python
IRJET - Object Detection using Deep Learning with OpenCV and Python
 
120_SEM_Special_Topics.ppt
120_SEM_Special_Topics.ppt120_SEM_Special_Topics.ppt
120_SEM_Special_Topics.ppt
 
poster09
poster09poster09
poster09
 
Sgg crest-presentation-final
Sgg crest-presentation-finalSgg crest-presentation-final
Sgg crest-presentation-final
 
Deep Learning for X ray Image to Text Generation
Deep Learning for X ray Image to Text GenerationDeep Learning for X ray Image to Text Generation
Deep Learning for X ray Image to Text Generation
 
IRJET- Deep Convolution Neural Networks for Galaxy Morphology Classification
IRJET- Deep Convolution Neural Networks for Galaxy Morphology ClassificationIRJET- Deep Convolution Neural Networks for Galaxy Morphology Classification
IRJET- Deep Convolution Neural Networks for Galaxy Morphology Classification
 
Emc 2013 Big Data in Astronomy
Emc 2013 Big Data in AstronomyEmc 2013 Big Data in Astronomy
Emc 2013 Big Data in Astronomy
 
Optics group research overview
Optics group research overviewOptics group research overview
Optics group research overview
 
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...
Dorr Space Variant Spatio Temporal Filtering Of Video For Gaze Visualization ...
 
Comaskey_William_Poster_SULI_FALL_2014
Comaskey_William_Poster_SULI_FALL_2014Comaskey_William_Poster_SULI_FALL_2014
Comaskey_William_Poster_SULI_FALL_2014
 
Shadow Detection and Removal Techniques A Perspective View
Shadow Detection and Removal Techniques A Perspective ViewShadow Detection and Removal Techniques A Perspective View
Shadow Detection and Removal Techniques A Perspective View
 
What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care?
 
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million Galaxies
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million GalaxiesAstronomaly at Scale: Searching for Anomalies Amongst 4 Million Galaxies
Astronomaly at Scale: Searching for Anomalies Amongst 4 Million Galaxies
 
IRJET - Direct Me-Nevigation for Blind People
IRJET -  	  Direct Me-Nevigation for Blind PeopleIRJET -  	  Direct Me-Nevigation for Blind People
IRJET - Direct Me-Nevigation for Blind People
 
LLNL_poster_15.compressed
LLNL_poster_15.compressedLLNL_poster_15.compressed
LLNL_poster_15.compressed
 
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...
Social Network Analysis: What It Is, Why We Should Care, and What We Can Lear...
 
Machine Learning ICS 273A
Machine Learning ICS 273AMachine Learning ICS 273A
Machine Learning ICS 273A
 

More from Ian Mulvany

Mendeley and Activity Data
Mendeley and Activity DataMendeley and Activity Data
Mendeley and Activity DataIan Mulvany
 
Unveiling the web, making the implicit explicit.
Unveiling the web, making the implicit explicit.Unveiling the web, making the implicit explicit.
Unveiling the web, making the implicit explicit.Ian Mulvany
 
Telstar cambridge-2010-07-22-im.key
Telstar cambridge-2010-07-22-im.keyTelstar cambridge-2010-07-22-im.key
Telstar cambridge-2010-07-22-im.keyIan Mulvany
 
Growing Beyond Journals, Nature Web Applications
Growing Beyond Journals, Nature Web ApplicationsGrowing Beyond Journals, Nature Web Applications
Growing Beyond Journals, Nature Web ApplicationsIan Mulvany
 
Potential Of Technology
Potential Of TechnologyPotential Of Technology
Potential Of TechnologyIan Mulvany
 
A Cabinet Of Web2.0 Scientific Curiosities
A Cabinet Of Web2.0 Scientific CuriositiesA Cabinet Of Web2.0 Scientific Curiosities
A Cabinet Of Web2.0 Scientific CuriositiesIan Mulvany
 
Integrating Everyting
Integrating EverytingIntegrating Everyting
Integrating EverytingIan Mulvany
 
Manvsmachinewithnotes
ManvsmachinewithnotesManvsmachinewithnotes
ManvsmachinewithnotesIan Mulvany
 
Science and Web2.0
Science and Web2.0Science and Web2.0
Science and Web2.0Ian Mulvany
 
Digital Library Federation, Fall 07, Connotea Presentation
Digital Library Federation, Fall 07, Connotea PresentationDigital Library Federation, Fall 07, Connotea Presentation
Digital Library Federation, Fall 07, Connotea PresentationIan Mulvany
 
BarCamb Connotea by Ian Mulvany
BarCamb Connotea by Ian MulvanyBarCamb Connotea by Ian Mulvany
BarCamb Connotea by Ian MulvanyIan Mulvany
 

More from Ian Mulvany (12)

Mendeley and Activity Data
Mendeley and Activity DataMendeley and Activity Data
Mendeley and Activity Data
 
Unveiling the web, making the implicit explicit.
Unveiling the web, making the implicit explicit.Unveiling the web, making the implicit explicit.
Unveiling the web, making the implicit explicit.
 
Telstar cambridge-2010-07-22-im.key
Telstar cambridge-2010-07-22-im.keyTelstar cambridge-2010-07-22-im.key
Telstar cambridge-2010-07-22-im.key
 
Growing Beyond Journals, Nature Web Applications
Growing Beyond Journals, Nature Web ApplicationsGrowing Beyond Journals, Nature Web Applications
Growing Beyond Journals, Nature Web Applications
 
Potential Of Technology
Potential Of TechnologyPotential Of Technology
Potential Of Technology
 
A Cabinet Of Web2.0 Scientific Curiosities
A Cabinet Of Web2.0 Scientific CuriositiesA Cabinet Of Web2.0 Scientific Curiosities
A Cabinet Of Web2.0 Scientific Curiosities
 
Integrating Everyting
Integrating EverytingIntegrating Everyting
Integrating Everyting
 
Manvsmachine
ManvsmachineManvsmachine
Manvsmachine
 
Manvsmachinewithnotes
ManvsmachinewithnotesManvsmachinewithnotes
Manvsmachinewithnotes
 
Science and Web2.0
Science and Web2.0Science and Web2.0
Science and Web2.0
 
Digital Library Federation, Fall 07, Connotea Presentation
Digital Library Federation, Fall 07, Connotea PresentationDigital Library Federation, Fall 07, Connotea Presentation
Digital Library Federation, Fall 07, Connotea Presentation
 
BarCamb Connotea by Ian Mulvany
BarCamb Connotea by Ian MulvanyBarCamb Connotea by Ian Mulvany
BarCamb Connotea by Ian Mulvany
 

Recently uploaded

Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FMESafe Software
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonAnna Loughnan Colquhoun
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoffsammart93
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityPrincipled Technologies
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 
Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businesspanagenda
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsJoaquim Jorge
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationRadu Cotescu
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Miguel Araújo
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobeapidays
 
A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?Igalia
 
Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024The Digital Insurer
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...apidays
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...Martijn de Jong
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyKhushali Kathiriya
 

Recently uploaded (20)

Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers:  A Deep Dive into Serverless Spatial Data and FMECloud Frontiers:  A Deep Dive into Serverless Spatial Data and FME
Cloud Frontiers: A Deep Dive into Serverless Spatial Data and FME
 
Data Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt RobisonData Cloud, More than a CDP by Matt Robison
Data Cloud, More than a CDP by Matt Robison
 
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot TakeoffStrategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
Strategize a Smooth Tenant-to-tenant Migration and Copilot Takeoff
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 
Why Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire businessWhy Teams call analytics are critical to your entire business
Why Teams call analytics are critical to your entire business
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and Myths
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Scaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organizationScaling API-first – The story of a global engineering organization
Scaling API-first – The story of a global engineering organization
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?A Year of the Servo Reboot: Where Are We Now?
A Year of the Servo Reboot: Where Are We Now?
 
Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024Manulife - Insurer Innovation Award 2024
Manulife - Insurer Innovation Award 2024
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
Apidays New York 2024 - The Good, the Bad and the Governed by David O'Neill, ...
 
2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...2024: Domino Containers - The Next Step. News from the Domino Container commu...
2024: Domino Containers - The Next Step. News from the Domino Container commu...
 
Artificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : UncertaintyArtificial Intelligence Chap.5 : Uncertainty
Artificial Intelligence Chap.5 : Uncertainty
 

Mining Surprise

  • 1. Surprise, You’ve Got Friends! Problem Definition: Want to present the serendipitous connections? How do we define and mine surprise? Then Aspects of Surprise: Uncanny vs Unexpected Surprise can be measured as the change in prior to posterior probability, is time dependent. Approaches/Main Idea Explicit communities, e.g. flickr group, are also connected by implicit groups, e.g. personomy measures, co-location via geo tags. Can find two people at this moment, this place, with the greatest Explicit Distance, and smallest Implicit Distance, this may surprise or abhor these people. Now Weighting Consider geo similarity. Inverse weight geo location by radius of bounding box, or “fame” of location. Test Case Use geo tagged flickr pictures to find two strangers at a gathering that have a visited the same unusual location in the past.
  • 2. A unit of surprise --- a wow may be defined for a single model M as the amount of surprise corresponding to a two-fold variation between P(M|D) and P(M), i.e., as log P(M|D)/P(M) (with log taken in base 2) • L. Itti & P. Baldi, Proc. IEEE-CVPR, 2005 • L. Itti & P. Baldi, Proc. NIPS, 2006
  • 3.
  • 4. First approximation Took the flickr group ISWC      1.  Extract the tag distribution of the event based on the photos in the group(E) 2. For each user: Took the user vector and computer the TF-IDF using the group factor as the IDF 3. For each pair of users Compute the cosine similarity
  • 5. Results The two “Closest - furthest” people in this group are from the same research institute in Ireland, DERI
  • 6. Base Case The base-case, using all the information, gives the same result, however the fist pass analysis was naive
  • 7. Plenty of signals that we wanted to take into account: Network of friends (two individuals were friends in flickr) All tags from all users Geo Coordinates etc etc etc
  • 9. Some Broader Issues Surprise is time dependent, current signals from some SNS are highly bursty, e.g. batch upload in flickr Connecting this analysis to real world sensors would seem to offer a method for extracting better information on calculating priors. Connecting across data sources certainly will A Pleasure Device “Surprise Engine” may become a tunable mess, much like current search engines.