SlideShare una empresa de Scribd logo
1 de 23
Computing in Social Networks:
Building Recommendation
Systems on Social Data
Outlook
Introduction
Recommender Systems
Examples of recommender systems
Challenges with recommendation research
Social networks and recommendations
Show case of experimental work on:
Trust-aware recommendations
Privacy preserving recommendations
Diversity and opinions
Conclusion
NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
2
Personalization and recommendations
Problem:
• Information overload…
Personalization and Profiles
• Users want to get personalized experience and at the same time don’t
want to share a lot of their personal information.
Recommendation systems
• Referred to as a range of algorithms which suggest a collection of items
to users, based on the knowledge of their profiles or previous
interactions.
Recommendation systems types:
• Collaborative filtering (User-based)
• Content-based filtering (Item-based)
• Hybrid filtering (Mix of users and content)
NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
3
NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
4
Applications of
Recommendation Systems
Important Challenges in Recommendation
Research
1. Explaining the recommendations
It increases the trust of users as they know what is the basis
of the suggestions
2. Preserving the user privacy
How to make good recommendations without ignoring user
privacy
3. Diversity and novelty of recommendations
Recommenders suggest similar stuff to what you have seen,
it is important to get
5NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Social networks
Social networks [Wasserman et al,
1994]
• Focus of fields such as
behavioral, marketing, economics,
etc.
Relationships types
• Interactions, social relations
Explicit relationships
• Relations in online social networks
like in facebook, linkedin, etc).
Implicit relationships
• Computed based on users
behavior. For instance rating
movies, music, etc.
6NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Image: https://www.facebook.com/notes/facebook-
engineering/visualizing-friendships/
Benefits of using social networks
for recommendations
• Take advantage of social network structure:
• Trust, social and structural Influence, transitivity, etc.
• Resilient against fraud, spam and fake accounts
• Identity and connections of the people on a social
network helps on dealing with bad guys
• Cold start problem
• System always has people to suggest (as long as they
are connected to the social network)
7NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Experimental work with trust and
recommendations
• Extracting trust networks from
• Getting better reach to items and users for improved
guessing of items to suggest.
• Using trust (networks) to improve accuracy of
recommendations generated
• Accurate suggestions of movies to watch, people to
follow, etc.
8NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Visualization of Trust Relations in
Ciao Dataset
9NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
In Nima Dokoohaki, Shahab Mokarizadeh, Mihhail Matskin, Ramona Bunea.
Correlating Trust and Privacy in Recommender Systems,
Special Issue on Web Intelligence and Personalization on Social Media,
Web Intelligence and Agent Systems An International Journal. IOS Press, 2014.
(submitted for review)
Trust networks and recommendations:
Data: Ratings Profiles to Trust Networks
10NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Trust networks and recommendations:
Impact of Trust Metric on Generated
Networks Structure
11NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Generated Trust Networks for Top-10 Trustworthy Users
(n= 5, m= 5): Without T-index
Generated Trust Networks for Top-10 Trustworthy
Users (n= 5, m=5): With T-index (= 100)
Soude Fazeli, Alireza Zarghami, Nima Dokoohaki, Mihhail Matskin,
Mechanizing Social Trust-Aware Recommenders with T-index Augmented Trustworthiness,
In proceedings of the 7th International Conference on Trust, Privacy & Security in Digital Business (Trustbus 2010)
Trust networks and recommendations:
Prediction accuracy against the variations
of Trustworthiness and Neighborhood size
12NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Soude Fazeli, Alireza Zarghami, Nima Dokoohaki, Mihhail Matskin,
Elevating Prediction Accuracy in Trust-aware Collaborative Filtering Recommenders through T-index Metric and TopTrustee lists,
In the Journal of Emerging Technologies in Web Intelligence (JETWI),
Special Issue On Web Personalization, Reputation and Recommender Systems, 2010.
Trust networks and recommendations
Rating Prediction Accuracy against network
(neighborhood) size
13NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Influence of search range on item coverage and prediction accuracy for Epinions dataset.
Stefan Magureanu, Nima Dokoohaki, Shahab Mokarizadeh, Mihhail Matskin,
Epidemic Trust-Based Recommender Systems ,
In proceedings of 2012 ASE/IEEE International Social Computing Conference
(SocialCom2012)
Experimental work with
Privacy and recommendations
• Proposing for software architectures that improve privacy of
recommendations
• How much data should the system use, can we control
this amount ?
• Can we use enough data and still get decent
suggestions ?
14NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Privacy and recommendations:
Component Architectures for Preserving
Privacy during Computations
15NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Nima Dokoohaki, Cihan Kaleli, Huseyin Polat and Mihhail Matskin,
Achieving Optimal Privacy in Trust-Aware Collaborative Filtering Recommender
Systems, The Second International Conference on Social Informatics (SocInfo 10)
Ramona Bunea, Shahab Mokarizadeh, Nima Dokoohaki and Mihhail Matskin,
Exploiting Dynamic Privacy in Socially Regularized Recommenders,
PinSoDa: Privacy in Social Data, in conjunction with the 11th IEEE International Conference on Data Mining (ICDM 2012)
NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
16
Privacy and recommendations:
Comparing performance of recommendations generated
Ramona Bunea, Shahab Mokarizadeh, Nima Dokoohaki and Mihhail Matskin,
Exploiting Dynamic Privacy in Socially Regularized Recommenders,
PinSoDa: Privacy in Social Data, in conjunction with the 11th IEEE International Conference on Data Mining (ICDM 2012)
Experimental work with diversity and
opinions recommendations
• How to diversify the recommendations
• What models can be proposed to give better summary
of reviews
• How to improve the recommendations of opinions in terms
of accuracy and scalability
• What models can be proposed to find more similar
people to read their Tweets.
17NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Data: From Review Profiles to Topic models
18NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Recommending Summarized Reviews:
Comparing Customer Ratings and estimated Sentiments
19NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
Ralf Krestel, Nima Dokoohaki
Diversifying Review Rankings, Special issue on Big Social Data Analytics,
Elsevier Journal of Neural Networks, 2014. Submitted for review.
Diversifying Summarized Reviews:
Comparing Recency of Summarization
Strategy Comparing LDA and LM
20NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
21
Recommending Tweets:
Visualizing variations of topics for #wikileaks
and #eurozone tweets, 2011
Extended results from: Nima Dokoohaki, Mihhail Matskin,
Mining Divergent Opinion Trust Networks through Latent Dirichlet Allocation,
In proceedings of 2012 IEEE/ACM International Conference on Social Network Analysis and Mining (ASONAM 2012)
Recommending Users:
Link Prediction on inferred trust relations,
tweets from 2009
22NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN
AUROC vs Number of Topics (Cosine)AUROC vs Number of Topics (KLD)
Extended results from: Nima Dokoohaki, Mihhail Matskin,
Mining Divergent Opinion Trust Networks through Latent Dirichlet Allocation,
In proceedings of 2012 IEEE/ACM International Conference on Social Network Analysis and Mining
(ASONAM 2012)
Conclusion
• This trail of research and education will continue under the
trends of data science and big data.
• KTH and other European institutions are planning to
design and offer study programs on data science and
analytics to students, hopefully very soon…
• Thank you!
24NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL
RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT,
STOCKHOLM, SWEDEN

Más contenido relacionado

Destacado

Social Recommender Systems
Social Recommender SystemsSocial Recommender Systems
Social Recommender Systemsguest77b0cd12
 
Google Tech Talk on Social Recommendation
Google Tech Talk on Social RecommendationGoogle Tech Talk on Social Recommendation
Google Tech Talk on Social RecommendationDan Carroll
 
TurKit: A Toolkit for Human Computation Algorithms
TurKit: A Toolkit for Human Computation AlgorithmsTurKit: A Toolkit for Human Computation Algorithms
TurKit: A Toolkit for Human Computation AlgorithmsGreg Little
 
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...Search Computing
 
Fundchange and Koodonation Workshop Slides - Nov 23, 2011
Fundchange and Koodonation Workshop Slides - Nov 23, 2011Fundchange and Koodonation Workshop Slides - Nov 23, 2011
Fundchange and Koodonation Workshop Slides - Nov 23, 2011Ideavibes | Paul Dombowsky
 
Volunteer Anywhere
Volunteer AnywhereVolunteer Anywhere
Volunteer AnywhereHelpFromHome
 
Social media recommendation based on people and tags (final)
Social media recommendation based on people and tags (final)Social media recommendation based on people and tags (final)
Social media recommendation based on people and tags (final)es712
 
Social Recommendation
Social RecommendationSocial Recommendation
Social Recommendationgu wendong
 
RSWEB 2013: A research platform for social recommendation
RSWEB 2013: A research platform for social recommendationRSWEB 2013: A research platform for social recommendation
RSWEB 2013: A research platform for social recommendationAmit Sharma
 
Social networks security risks
Social networks security risksSocial networks security risks
Social networks security risksosuhaibany
 
Building Social Networks
Building Social NetworksBuilding Social Networks
Building Social Networksnyccamp
 
Introduction to cryptography
Introduction to cryptographyIntroduction to cryptography
Introduction to cryptographyAmir Neziri
 
Social Networks and Security: What Your Teenager Likely Won't Tell You
Social Networks and Security: What Your Teenager Likely Won't Tell YouSocial Networks and Security: What Your Teenager Likely Won't Tell You
Social Networks and Security: What Your Teenager Likely Won't Tell YouDenim Group
 
It Only Takes a Minute
It Only Takes a MinuteIt Only Takes a Minute
It Only Takes a Minuteelliottofhook
 
Data Visualization and Social Network Analysis for Recruiting.
Data Visualization and Social Network Analysis for Recruiting.Data Visualization and Social Network Analysis for Recruiting.
Data Visualization and Social Network Analysis for Recruiting.Matt Charney
 
Introduction to Cryptography Part I
Introduction to Cryptography Part IIntroduction to Cryptography Part I
Introduction to Cryptography Part IMaksim Djackov
 
FITC - Bootstrap Unleashed
FITC - Bootstrap UnleashedFITC - Bootstrap Unleashed
FITC - Bootstrap UnleashedRami Sayar
 
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)arborwebsolutions
 
Social journalism: Community building through social networks
Social journalism: Community building through social networksSocial journalism: Community building through social networks
Social journalism: Community building through social networksJD Lasica
 
Responsive Web Design - Introduction & Workflow Overview
Responsive Web Design - Introduction & Workflow OverviewResponsive Web Design - Introduction & Workflow Overview
Responsive Web Design - Introduction & Workflow OverviewAidan Foster
 

Destacado (20)

Social Recommender Systems
Social Recommender SystemsSocial Recommender Systems
Social Recommender Systems
 
Google Tech Talk on Social Recommendation
Google Tech Talk on Social RecommendationGoogle Tech Talk on Social Recommendation
Google Tech Talk on Social Recommendation
 
TurKit: A Toolkit for Human Computation Algorithms
TurKit: A Toolkit for Human Computation AlgorithmsTurKit: A Toolkit for Human Computation Algorithms
TurKit: A Toolkit for Human Computation Algorithms
 
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...
CrowdSearcher. Reactive and multiplatform Crowdsourcing. keynote speech at DB...
 
Fundchange and Koodonation Workshop Slides - Nov 23, 2011
Fundchange and Koodonation Workshop Slides - Nov 23, 2011Fundchange and Koodonation Workshop Slides - Nov 23, 2011
Fundchange and Koodonation Workshop Slides - Nov 23, 2011
 
Volunteer Anywhere
Volunteer AnywhereVolunteer Anywhere
Volunteer Anywhere
 
Social media recommendation based on people and tags (final)
Social media recommendation based on people and tags (final)Social media recommendation based on people and tags (final)
Social media recommendation based on people and tags (final)
 
Social Recommendation
Social RecommendationSocial Recommendation
Social Recommendation
 
RSWEB 2013: A research platform for social recommendation
RSWEB 2013: A research platform for social recommendationRSWEB 2013: A research platform for social recommendation
RSWEB 2013: A research platform for social recommendation
 
Social networks security risks
Social networks security risksSocial networks security risks
Social networks security risks
 
Building Social Networks
Building Social NetworksBuilding Social Networks
Building Social Networks
 
Introduction to cryptography
Introduction to cryptographyIntroduction to cryptography
Introduction to cryptography
 
Social Networks and Security: What Your Teenager Likely Won't Tell You
Social Networks and Security: What Your Teenager Likely Won't Tell YouSocial Networks and Security: What Your Teenager Likely Won't Tell You
Social Networks and Security: What Your Teenager Likely Won't Tell You
 
It Only Takes a Minute
It Only Takes a MinuteIt Only Takes a Minute
It Only Takes a Minute
 
Data Visualization and Social Network Analysis for Recruiting.
Data Visualization and Social Network Analysis for Recruiting.Data Visualization and Social Network Analysis for Recruiting.
Data Visualization and Social Network Analysis for Recruiting.
 
Introduction to Cryptography Part I
Introduction to Cryptography Part IIntroduction to Cryptography Part I
Introduction to Cryptography Part I
 
FITC - Bootstrap Unleashed
FITC - Bootstrap UnleashedFITC - Bootstrap Unleashed
FITC - Bootstrap Unleashed
 
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)
(Practical) Beyond Responsive Web Design (WordCamp Miami 2011)
 
Social journalism: Community building through social networks
Social journalism: Community building through social networksSocial journalism: Community building through social networks
Social journalism: Community building through social networks
 
Responsive Web Design - Introduction & Workflow Overview
Responsive Web Design - Introduction & Workflow OverviewResponsive Web Design - Introduction & Workflow Overview
Responsive Web Design - Introduction & Workflow Overview
 

Similar a Building Recommendation Systems on Social Data @KTH - FutureFriday - March 2014

Short CfP #DISC2016
Short CfP #DISC2016Short CfP #DISC2016
Short CfP #DISC2016Han Woo PARK
 
Final call for #DISC2016
Final call for #DISC2016Final call for #DISC2016
Final call for #DISC2016Kyujin Jung
 
Distributed Trust Architecture: The New Foundation of Everything
Distributed Trust Architecture: The New Foundation of EverythingDistributed Trust Architecture: The New Foundation of Everything
Distributed Trust Architecture: The New Foundation of EverythingLiming Zhu
 
Big social data analytics - social network analysis
Big social data analytics - social network analysis Big social data analytics - social network analysis
Big social data analytics - social network analysis Jari Jussila
 
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...Florence Hudson
 
Values Interventions: Ethics Scholarship in Action
Values Interventions: Ethics Scholarship in ActionValues Interventions: Ethics Scholarship in Action
Values Interventions: Ethics Scholarship in ActionMichael Zimmer
 
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-decke-SIDES.eu
 
2015-10-14 research seminar 2
2015-10-14 research seminar 22015-10-14 research seminar 2
2015-10-14 research seminar 2ifi8106tlu
 
Influencing the MOOC agenda - analysis of #MOOC Twitter Data
Influencing the MOOC agenda - analysis of #MOOC Twitter Data  Influencing the MOOC agenda - analysis of #MOOC Twitter Data
Influencing the MOOC agenda - analysis of #MOOC Twitter Data Mairéad Nic Giolla Mhichíl
 
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ Project
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ ProjectQuantum Mechanics meet Information Search and Retrieval – The QUARTZ Project
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ ProjectIngo Frommholz
 
ESRC Research Methods Festival NSMNSS presentation
ESRC Research Methods Festival NSMNSS presentationESRC Research Methods Festival NSMNSS presentation
ESRC Research Methods Festival NSMNSS presentationKandy Woodfield
 
Computational methods for intelligent matchmaking for knowledge work
Computational methods for intelligent matchmaking for knowledge workComputational methods for intelligent matchmaking for knowledge work
Computational methods for intelligent matchmaking for knowledge workJari Jussila
 
Global Computing: an Analysis of Trust and Wireless Communications
Global Computing: an Analysis of Trust and Wireless CommunicationsGlobal Computing: an Analysis of Trust and Wireless Communications
Global Computing: an Analysis of Trust and Wireless CommunicationsNicola Mezzetti
 
The challenges and benefits of using digital to engage people in research
The challenges and benefits of using digital to engage people in researchThe challenges and benefits of using digital to engage people in research
The challenges and benefits of using digital to engage people in researchKirsten Thompson
 
Social Media Research at Comms Service Providers
Social Media Research at Comms Service ProvidersSocial Media Research at Comms Service Providers
Social Media Research at Comms Service ProvidersDavid Strom
 
acatech_STUDY_Internet_Privacy_WEB
acatech_STUDY_Internet_Privacy_WEBacatech_STUDY_Internet_Privacy_WEB
acatech_STUDY_Internet_Privacy_WEBJaina Hirai
 
Market Research in the Mobile World 2010
Market Research in the Mobile World 2010Market Research in the Mobile World 2010
Market Research in the Mobile World 2010Merlien Institute
 
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...Cataldo Musto
 

Similar a Building Recommendation Systems on Social Data @KTH - FutureFriday - March 2014 (20)

Short CfP #DISC2016
Short CfP #DISC2016Short CfP #DISC2016
Short CfP #DISC2016
 
Final call for #DISC2016
Final call for #DISC2016Final call for #DISC2016
Final call for #DISC2016
 
Distributed Trust Architecture: The New Foundation of Everything
Distributed Trust Architecture: The New Foundation of EverythingDistributed Trust Architecture: The New Foundation of Everything
Distributed Trust Architecture: The New Foundation of Everything
 
Big social data analytics - social network analysis
Big social data analytics - social network analysis Big social data analytics - social network analysis
Big social data analytics - social network analysis
 
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...
PEARC17: ARCC Identity and Access Management, Security and related topics. Cy...
 
Values Interventions: Ethics Scholarship in Action
Values Interventions: Ethics Scholarship in ActionValues Interventions: Ethics Scholarship in Action
Values Interventions: Ethics Scholarship in Action
 
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck
"Towards Value-Centric Big Data" e-SIDES Workshop - Slide-deck
 
2015-10-14 research seminar 2
2015-10-14 research seminar 22015-10-14 research seminar 2
2015-10-14 research seminar 2
 
Influencing the MOOC agenda - analysis of #MOOC Twitter Data
Influencing the MOOC agenda - analysis of #MOOC Twitter Data  Influencing the MOOC agenda - analysis of #MOOC Twitter Data
Influencing the MOOC agenda - analysis of #MOOC Twitter Data
 
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ Project
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ ProjectQuantum Mechanics meet Information Search and Retrieval – The QUARTZ Project
Quantum Mechanics meet Information Search and Retrieval – The QUARTZ Project
 
ESRC Research Methods Festival NSMNSS presentation
ESRC Research Methods Festival NSMNSS presentationESRC Research Methods Festival NSMNSS presentation
ESRC Research Methods Festival NSMNSS presentation
 
Computational methods for intelligent matchmaking for knowledge work
Computational methods for intelligent matchmaking for knowledge workComputational methods for intelligent matchmaking for knowledge work
Computational methods for intelligent matchmaking for knowledge work
 
Global Computing: an Analysis of Trust and Wireless Communications
Global Computing: an Analysis of Trust and Wireless CommunicationsGlobal Computing: an Analysis of Trust and Wireless Communications
Global Computing: an Analysis of Trust and Wireless Communications
 
The challenges and benefits of using digital to engage people in research
The challenges and benefits of using digital to engage people in researchThe challenges and benefits of using digital to engage people in research
The challenges and benefits of using digital to engage people in research
 
Social Media Research at Comms Service Providers
Social Media Research at Comms Service ProvidersSocial Media Research at Comms Service Providers
Social Media Research at Comms Service Providers
 
Cook social network innovation
Cook   social network innovationCook   social network innovation
Cook social network innovation
 
acatech_STUDY_Internet_Privacy_WEB
acatech_STUDY_Internet_Privacy_WEBacatech_STUDY_Internet_Privacy_WEB
acatech_STUDY_Internet_Privacy_WEB
 
Market Research in the Mobile World 2010
Market Research in the Mobile World 2010Market Research in the Mobile World 2010
Market Research in the Mobile World 2010
 
Sport && Wellness Hackathon
Sport && Wellness  HackathonSport && Wellness  Hackathon
Sport && Wellness Hackathon
 
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...
DeCAT 2015 - International Workshop on Deep Content Analytics Techniques for ...
 

Último

AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAndrey Devyatkin
 
Exploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusExploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusZilliz
 
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ..."I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...Zilliz
 
Platformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityPlatformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityWSO2
 
DBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor PresentationDBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor PresentationDropbox
 
WSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MIND CTI
 
CNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In PakistanCNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In Pakistandanishmna97
 
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024Victor Rentea
 
FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024The Digital Insurer
 
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
 
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
 
ICT role in 21st century education and its challenges
ICT role in 21st century education and its challengesICT role in 21st century education and its challenges
ICT role in 21st century education and its challengesrafiqahmad00786416
 
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...apidays
 
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Orbitshub
 
Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​Bhuvaneswari Subramani
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdfSandro Moreira
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWERMadyBayot
 
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfRising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfOrbitshub
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 

Último (20)

AWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of TerraformAWS Community Day CPH - Three problems of Terraform
AWS Community Day CPH - Three problems of Terraform
 
Exploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with MilvusExploring Multimodal Embeddings with Milvus
Exploring Multimodal Embeddings with Milvus
 
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ..."I see eyes in my soup": How Delivery Hero implemented the safety system for ...
"I see eyes in my soup": How Delivery Hero implemented the safety system for ...
 
Platformless Horizons for Digital Adaptability
Platformless Horizons for Digital AdaptabilityPlatformless Horizons for Digital Adaptability
Platformless Horizons for Digital Adaptability
 
DBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor PresentationDBX First Quarter 2024 Investor Presentation
DBX First Quarter 2024 Investor Presentation
 
WSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering DevelopersWSO2's API Vision: Unifying Control, Empowering Developers
WSO2's API Vision: Unifying Control, Empowering Developers
 
MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024MINDCTI Revenue Release Quarter One 2024
MINDCTI Revenue Release Quarter One 2024
 
CNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In PakistanCNIC Information System with Pakdata Cf In Pakistan
CNIC Information System with Pakdata Cf In Pakistan
 
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024
Modular Monolith - a Practical Alternative to Microservices @ Devoxx UK 2024
 
FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024FWD Group - Insurer Innovation Award 2024
FWD Group - Insurer Innovation Award 2024
 
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, ...
 
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
 
ICT role in 21st century education and its challenges
ICT role in 21st century education and its challengesICT role in 21st century education and its challenges
ICT role in 21st century education and its challenges
 
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
Apidays New York 2024 - Accelerating FinTech Innovation by Vasa Krishnan, Fin...
 
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
Navigating the Deluge_ Dubai Floods and the Resilience of Dubai International...
 
Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​Elevate Developer Efficiency & build GenAI Application with Amazon Q​
Elevate Developer Efficiency & build GenAI Application with Amazon Q​
 
[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf[BuildWithAI] Introduction to Gemini.pdf
[BuildWithAI] Introduction to Gemini.pdf
 
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWEREMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
EMPOWERMENT TECHNOLOGY GRADE 11 QUARTER 2 REVIEWER
 
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdfRising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
Rising Above_ Dubai Floods and the Fortitude of Dubai International Airport.pdf
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 

Building Recommendation Systems on Social Data @KTH - FutureFriday - March 2014

  • 1. Computing in Social Networks: Building Recommendation Systems on Social Data
  • 2. Outlook Introduction Recommender Systems Examples of recommender systems Challenges with recommendation research Social networks and recommendations Show case of experimental work on: Trust-aware recommendations Privacy preserving recommendations Diversity and opinions Conclusion NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN 2
  • 3. Personalization and recommendations Problem: • Information overload… Personalization and Profiles • Users want to get personalized experience and at the same time don’t want to share a lot of their personal information. Recommendation systems • Referred to as a range of algorithms which suggest a collection of items to users, based on the knowledge of their profiles or previous interactions. Recommendation systems types: • Collaborative filtering (User-based) • Content-based filtering (Item-based) • Hybrid filtering (Mix of users and content) NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN 3
  • 4. NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN 4 Applications of Recommendation Systems
  • 5. Important Challenges in Recommendation Research 1. Explaining the recommendations It increases the trust of users as they know what is the basis of the suggestions 2. Preserving the user privacy How to make good recommendations without ignoring user privacy 3. Diversity and novelty of recommendations Recommenders suggest similar stuff to what you have seen, it is important to get 5NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 6. Social networks Social networks [Wasserman et al, 1994] • Focus of fields such as behavioral, marketing, economics, etc. Relationships types • Interactions, social relations Explicit relationships • Relations in online social networks like in facebook, linkedin, etc). Implicit relationships • Computed based on users behavior. For instance rating movies, music, etc. 6NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Image: https://www.facebook.com/notes/facebook- engineering/visualizing-friendships/
  • 7. Benefits of using social networks for recommendations • Take advantage of social network structure: • Trust, social and structural Influence, transitivity, etc. • Resilient against fraud, spam and fake accounts • Identity and connections of the people on a social network helps on dealing with bad guys • Cold start problem • System always has people to suggest (as long as they are connected to the social network) 7NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 8. Experimental work with trust and recommendations • Extracting trust networks from • Getting better reach to items and users for improved guessing of items to suggest. • Using trust (networks) to improve accuracy of recommendations generated • Accurate suggestions of movies to watch, people to follow, etc. 8NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 9. Visualization of Trust Relations in Ciao Dataset 9NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN In Nima Dokoohaki, Shahab Mokarizadeh, Mihhail Matskin, Ramona Bunea. Correlating Trust and Privacy in Recommender Systems, Special Issue on Web Intelligence and Personalization on Social Media, Web Intelligence and Agent Systems An International Journal. IOS Press, 2014. (submitted for review)
  • 10. Trust networks and recommendations: Data: Ratings Profiles to Trust Networks 10NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 11. Trust networks and recommendations: Impact of Trust Metric on Generated Networks Structure 11NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Generated Trust Networks for Top-10 Trustworthy Users (n= 5, m= 5): Without T-index Generated Trust Networks for Top-10 Trustworthy Users (n= 5, m=5): With T-index (= 100) Soude Fazeli, Alireza Zarghami, Nima Dokoohaki, Mihhail Matskin, Mechanizing Social Trust-Aware Recommenders with T-index Augmented Trustworthiness, In proceedings of the 7th International Conference on Trust, Privacy & Security in Digital Business (Trustbus 2010)
  • 12. Trust networks and recommendations: Prediction accuracy against the variations of Trustworthiness and Neighborhood size 12NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Soude Fazeli, Alireza Zarghami, Nima Dokoohaki, Mihhail Matskin, Elevating Prediction Accuracy in Trust-aware Collaborative Filtering Recommenders through T-index Metric and TopTrustee lists, In the Journal of Emerging Technologies in Web Intelligence (JETWI), Special Issue On Web Personalization, Reputation and Recommender Systems, 2010.
  • 13. Trust networks and recommendations Rating Prediction Accuracy against network (neighborhood) size 13NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Influence of search range on item coverage and prediction accuracy for Epinions dataset. Stefan Magureanu, Nima Dokoohaki, Shahab Mokarizadeh, Mihhail Matskin, Epidemic Trust-Based Recommender Systems , In proceedings of 2012 ASE/IEEE International Social Computing Conference (SocialCom2012)
  • 14. Experimental work with Privacy and recommendations • Proposing for software architectures that improve privacy of recommendations • How much data should the system use, can we control this amount ? • Can we use enough data and still get decent suggestions ? 14NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 15. Privacy and recommendations: Component Architectures for Preserving Privacy during Computations 15NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Nima Dokoohaki, Cihan Kaleli, Huseyin Polat and Mihhail Matskin, Achieving Optimal Privacy in Trust-Aware Collaborative Filtering Recommender Systems, The Second International Conference on Social Informatics (SocInfo 10) Ramona Bunea, Shahab Mokarizadeh, Nima Dokoohaki and Mihhail Matskin, Exploiting Dynamic Privacy in Socially Regularized Recommenders, PinSoDa: Privacy in Social Data, in conjunction with the 11th IEEE International Conference on Data Mining (ICDM 2012)
  • 16. NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN 16 Privacy and recommendations: Comparing performance of recommendations generated Ramona Bunea, Shahab Mokarizadeh, Nima Dokoohaki and Mihhail Matskin, Exploiting Dynamic Privacy in Socially Regularized Recommenders, PinSoDa: Privacy in Social Data, in conjunction with the 11th IEEE International Conference on Data Mining (ICDM 2012)
  • 17. Experimental work with diversity and opinions recommendations • How to diversify the recommendations • What models can be proposed to give better summary of reviews • How to improve the recommendations of opinions in terms of accuracy and scalability • What models can be proposed to find more similar people to read their Tweets. 17NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 18. Data: From Review Profiles to Topic models 18NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 19. Recommending Summarized Reviews: Comparing Customer Ratings and estimated Sentiments 19NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN Ralf Krestel, Nima Dokoohaki Diversifying Review Rankings, Special issue on Big Social Data Analytics, Elsevier Journal of Neural Networks, 2014. Submitted for review.
  • 20. Diversifying Summarized Reviews: Comparing Recency of Summarization Strategy Comparing LDA and LM 20NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN
  • 21. NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN 21 Recommending Tweets: Visualizing variations of topics for #wikileaks and #eurozone tweets, 2011 Extended results from: Nima Dokoohaki, Mihhail Matskin, Mining Divergent Opinion Trust Networks through Latent Dirichlet Allocation, In proceedings of 2012 IEEE/ACM International Conference on Social Network Analysis and Mining (ASONAM 2012)
  • 22. Recommending Users: Link Prediction on inferred trust relations, tweets from 2009 22NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN AUROC vs Number of Topics (Cosine)AUROC vs Number of Topics (KLD) Extended results from: Nima Dokoohaki, Mihhail Matskin, Mining Divergent Opinion Trust Networks through Latent Dirichlet Allocation, In proceedings of 2012 IEEE/ACM International Conference on Social Network Analysis and Mining (ASONAM 2012)
  • 23. Conclusion • This trail of research and education will continue under the trends of data science and big data. • KTH and other European institutions are planning to design and offer study programs on data science and analytics to students, hopefully very soon… • Thank you! 24NIMA DOKOOHAKI, NIMAD@KTH.SE POSTDOCTORAL RESEARCHER SEMINAR @ FUTURE FRIDAY 2014 KTH/ICT, STOCKHOLM, SWEDEN

Notas del editor

  1. 30 minutes Would be 15 slides + 5 minutes questions I guess Image CC: https://www.flickr.com/photos/daniel_iversen/5440728466/sizes/m/in/photostream/
  2. https://www.flickr.com/photos/daviderickson/5579493777/sizes/o/in/photostream/ https://www.flickr.com/photos/daviderickson/5580079906/sizes/o/in/photostream/ https://www.flickr.com/photos/42696116@N00/3979783546/in/photoli https://www.flickr.com/photos/stevegarfield/
  3. Social network of Google+ Image: http://www.flickr.com/photos/ajc1/6260304760/
  4. For the sake of simplicity, we display only users(displayed as nodes) and their connections (trust relationships) to top-10 trustworthy users. As mentioned, each cluster is described as a group of like-minded users in terms of trust. It is shown that the number of common users between clusters increases which enables users of different clusters to find each other easier. In our case, more users form divergent areas of users’ interests, presented as clusters, can be accessible.
  5. Results have been partially competable with Neil Lathia’s work
  6. ROC curve for Pearson (left) and Kullback-Leibler (right) Variable: social network size