SlideShare ist ein Scribd-Unternehmen logo
1 von 20
Downloaden Sie, um offline zu lesen
Exploring Generative Models of Tripartite
Graphs for Recommendation in Social Media
Charalampos Chelmis, Viktor K Prasanna
chelmis@usc.edu
MSM 2013, Paris, France
• Introduction
• Structure of Tripartite Graphs
• Generative Models of Tripartite Graphs
• Social Link Classification Schemes
• Evaluation
• Conclusion
Overview
2
• Social Networking is used for
 Content organization
 Content sharing
• Multiple media types
• Users' activities
 Reveal interests and tastes
 Hidden structure
• Description of Resources
 Text
 Tags / Hashtags
• Social Annotation
 Collective characterization of resources
 Use of synonyms for similar recourses
 Same keywords for different recourses
Introduction
3
• How to address issues of synonymy and polysemy?
 Deal with space size explosion
• How to discover emergent structure in online tagging systems?
 Hidden topics
• How to capture users’ latent interests?
 Which subjects a user is mostly interested in?
 Which users have similar interests?
• How to model the process of social generation of annotations?
 How to capture the semantics of collaboration
• Why is this useful?
 Recommend people
 Recommend Tags / resources
 Clustering
 …
Research Questions
4
• Set of actors (e.g. users) A={a1, ...,ak}
• Set of concepts (e.g. tags) C = {c1, ..., cl}
• Set of resources (e.g. photos) R ={r1, ..., rm}
Structure of Tripartite Graphs
5
• The User-Concept Model
 Users are modeled based on their tag usage
 φ denotes the matrix of topic distributions
− multinomial distribution over N concepts
− T topics being drawn independently
 θ: the matrix of user-specific mixture weights for
these T topics
• Captures users’ latent interests
• Ignores Resources
• Ignores the social aspect of tagging
• The User-Resource Model
 Resources become vocabulary terms
• Tags are ignored
• Ignores the social aspect of tagging
Reducing the Tripartite Graph to Bipartite Structures
6
• Topic-based representation
• Model both resources & users’ interests
• Multiple users may annotate resource r
• For each tag a user is chosen uniformly at random
• Each user is associated with a distribution over
latent topics ɵ
• A topic is chosen from a distribution over topics
specific to that user
• The tag is generated from the chosen topic
 φt: probability distribution of tags for topic t
The User-Resource-Concept Model
7
• Tag Recommendation
 Automatic annotation enhancement
 Search improvement
• Clustering
 Community detection
 Organization of resources/tags in categories
• Navigation and Visualization
 Social browsing
• Next we focus on recommending people
Recommendation
8
• Classification Based on Latent Interests
 Measure “tastes” distance with respect to latent topics distribution
 Pointwise squared distance between feature vectors of users u and v

 Other measures to consider
− Kullback Leibler (KL) divergence
− Cosine similarity
• Objective:
 Minimize the distance between linked users
• Focus on topical homophily
 Ignore network effects
• Prior work uses network proximity as indicator of link formation
Social Link Recommendation Using
Latent Semantics & Network Structure
9
]v))(k,-u)(k,(,,v))(1,-u)(1,[(v)F(u, 22
ΘΘΘΘ= 
F(u,v) = 0 => u,v have
identical distributions
F(u,v) > 0 => distributions
diverge
• Latent Topics & Local Structure
 CN(u,v) = common neighbors between users u and v
− Simplicity and computational efficiency
 Latent topics similarity


• Latent Topics & Global Structure
 SD(u,v) = shortest distance between users u and v

• Non separable training set => inefficient classifiers
• Aggregation Strategy
 Reduce the number of training samples
 Produce more efficient classifiers
 Average latent similarity of user pairs with k common
neighbors:
Social Link Recommendation Using
Latent Semantics & Network Structure
10
v)]CN(u,v),(u,[v)F(u, σ=
∑==
=
kk:pp p
(p)
|kk:p|
1
(k)avg σσ
v)]SD(u,v),(u,[v)F(u, σ=
22
),(),(
),(),(
),(
∑∑
∑
ΘΘ
ΘΘ
=
tt
t
vtut
vtut
vuσ
• Objectives
 Ability to uncover subliminal collective knowledge
 Evaluate performance of “people” recommendation
• Setting
 2.4 GHz Intel Core 2 Duo, 2 GB memory, Windows 7
• Real-world Dataset
 Last.fm online music system
− social relationships
− tagging information
− music artist listening information
 Statistics
− 1,892 users
− 25,434 directed user friend relations
− 17,632 artists UR Model vocabulary size
− 92,834 user-listened-artist relations
− 11,946 unique tags UC and URC vocabulary size
− 186,479 annotations (tuples <user, tag, artist>)
Experimental Analysis
11
Sample Topics
12
• Evaluate ability to predict tags/resources on new users
 Perplexity
• Split dataset into two disjoint sets
 90% for training
• Lower perplexity indicates better generalization
• URC better overall
 Exploits more information
• UC
 Organizes tags in “clusters”
• UR
 Inferior quality due to noise
Predictive Power
13
• Split dataset into two disjoint sets
 10%, 25%, 50%, 75% for training, rest for testing
• Evaluation process
 Randomly sample 12,716 pairs of users
 50% true links, 50% negative samples
 Compute similarity of user pairs
 Sort users in decreasing order of similarity
 Add links between users with highest similarity
Recommendation of Social Ties
14
• Latent Topics & Shortest Distance
 Aggregates all true links training similarity values in a single point
 Least effective
• Ensemble achieves best precision
• Over fitting for training size > 50%
• Recall drops as dataset size increases
Recommendation of Social Ties
15
[Latent Topics & Local Structure]
[Latent Topics]
[Ensemble]
• In social media number of true links << absent links
• High performance for both classes
 True negatives easier to classify correctly
 Degradation in performance for true positives
• Reasonable results for practical purposes
How about High Class Imbalance?
16
[Latent Topics & Local Structure]
[Latent Topics]
[Ensemble]
• Baselines
 Cosine Similarity (CS)
 Maximal Information Path (MIP)
• Evaluation Criterion
 Area under the receiver-operating characteristic curve (AUC)
• Baselines AUC
 Computed over the complete dataset
 Biases the evaluation in favor of the baselines
 CS AUC = 0.6087
 MIP AUC = 0.6256
• Same evaluation process as before
• Compute performance lift
 % change over best performing baseline
 Positive % denotes improvement
Comparison to Tag-based similarity metrics
17
• Not all schemes can beat the baseline
 For 10% training data
 ≤10% AUC loss
 But, significant speedup due to minimal training dataset
• Latent Topics & Local Structure Scheme consistently better
Comparison to Tag-based similarity metrics
18
Training dataset size
[Latent Topics & Local Structure]
[Latent Topics]
• Three generative models of tripartite graphs in social tagging
systems
• Modeling of users’ interests in a latent space over resources and
metadata
• Limitations
 Ignore several aspects of real-world annotation process, such as topic
correlation and user interaction
• Achieve great performance in the recommendation task
 Accurate predictors of social ties in conjunction with structural
evidence
 Proposed aggregation strategy to reduce number of training samples
• Future work
 Incorporate other types of resources
 Automatically identify most discriminative latent topics and discard
uninformative resources and metadata
Concluding Remarks
19
• Questions?
chelmis@usc.edu
Thank you!
20

Weitere ähnliche Inhalte

Was ist angesagt?

Introduction to Topological Data Analysis
Introduction to Topological Data AnalysisIntroduction to Topological Data Analysis
Introduction to Topological Data AnalysisMason Porter
 
Social Network Analysis
Social Network AnalysisSocial Network Analysis
Social Network AnalysisSujoy Bag
 
04 Diffusion and Peer Influence
04 Diffusion and Peer Influence04 Diffusion and Peer Influence
04 Diffusion and Peer Influencednac
 
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
 
11 Network Experiments and Interventions
11 Network Experiments and Interventions11 Network Experiments and Interventions
11 Network Experiments and Interventionsdnac
 
10 More than a Pretty Picture: Visual Thinking in Network Studies
10 More than a Pretty Picture: Visual Thinking in Network Studies10 More than a Pretty Picture: Visual Thinking in Network Studies
10 More than a Pretty Picture: Visual Thinking in Network Studiesdnac
 
02 Network Data Collection
02 Network Data Collection02 Network Data Collection
02 Network Data Collectiondnac
 
01 Introduction to Networks Methods and Measures
01 Introduction to Networks Methods and Measures01 Introduction to Networks Methods and Measures
01 Introduction to Networks Methods and Measuresdnac
 
Topological Data Analysis of Complex Spatial Systems
Topological Data Analysis of Complex Spatial SystemsTopological Data Analysis of Complex Spatial Systems
Topological Data Analysis of Complex Spatial SystemsMason Porter
 
#lak2013, Leuven, DC slides, #learninganalytics
#lak2013, Leuven, DC slides, #learninganalytics#lak2013, Leuven, DC slides, #learninganalytics
#lak2013, Leuven, DC slides, #learninganalyticsSoudé Fazeli
 
Contextualized versus Structural Overlapping Communities in Social Media.
Contextualized versus Structural Overlapping Communities in Social Media. Contextualized versus Structural Overlapping Communities in Social Media.
Contextualized versus Structural Overlapping Communities in Social Media. Mohsen Shahriari
 
DeLiddo&BuckinghamShum-e-Part2014
DeLiddo&BuckinghamShum-e-Part2014DeLiddo&BuckinghamShum-e-Part2014
DeLiddo&BuckinghamShum-e-Part2014Anna De Liddo
 
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...Julita Vassileva
 
Centrality in Time- Dependent Networks
Centrality in Time- Dependent NetworksCentrality in Time- Dependent Networks
Centrality in Time- Dependent NetworksMason Porter
 
Paper Writing in Applied Mathematics (slightly updated slides)
Paper Writing in Applied Mathematics (slightly updated slides)Paper Writing in Applied Mathematics (slightly updated slides)
Paper Writing in Applied Mathematics (slightly updated slides)Mason Porter
 
Social Network Analysis (Part 1)
Social Network Analysis (Part 1)Social Network Analysis (Part 1)
Social Network Analysis (Part 1)Vala Ali Rohani
 
The Mathematics of Social Network Analysis: Metrics for Academic Social Networks
The Mathematics of Social Network Analysis: Metrics for Academic Social NetworksThe Mathematics of Social Network Analysis: Metrics for Academic Social Networks
The Mathematics of Social Network Analysis: Metrics for Academic Social NetworksEditor IJCATR
 
4C13 J.15 Larson "Twitter based discourse community"
4C13 J.15 Larson "Twitter based discourse community"4C13 J.15 Larson "Twitter based discourse community"
4C13 J.15 Larson "Twitter based discourse community"rhetoricked
 
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...GUANGYUAN PIAO
 

Was ist angesagt? (20)

Introduction to Topological Data Analysis
Introduction to Topological Data AnalysisIntroduction to Topological Data Analysis
Introduction to Topological Data Analysis
 
Social Network Analysis
Social Network AnalysisSocial Network Analysis
Social Network Analysis
 
DIE 20130724
DIE 20130724DIE 20130724
DIE 20130724
 
04 Diffusion and Peer Influence
04 Diffusion and Peer Influence04 Diffusion and Peer Influence
04 Diffusion and Peer Influence
 
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...
 
11 Network Experiments and Interventions
11 Network Experiments and Interventions11 Network Experiments and Interventions
11 Network Experiments and Interventions
 
10 More than a Pretty Picture: Visual Thinking in Network Studies
10 More than a Pretty Picture: Visual Thinking in Network Studies10 More than a Pretty Picture: Visual Thinking in Network Studies
10 More than a Pretty Picture: Visual Thinking in Network Studies
 
02 Network Data Collection
02 Network Data Collection02 Network Data Collection
02 Network Data Collection
 
01 Introduction to Networks Methods and Measures
01 Introduction to Networks Methods and Measures01 Introduction to Networks Methods and Measures
01 Introduction to Networks Methods and Measures
 
Topological Data Analysis of Complex Spatial Systems
Topological Data Analysis of Complex Spatial SystemsTopological Data Analysis of Complex Spatial Systems
Topological Data Analysis of Complex Spatial Systems
 
#lak2013, Leuven, DC slides, #learninganalytics
#lak2013, Leuven, DC slides, #learninganalytics#lak2013, Leuven, DC slides, #learninganalytics
#lak2013, Leuven, DC slides, #learninganalytics
 
Contextualized versus Structural Overlapping Communities in Social Media.
Contextualized versus Structural Overlapping Communities in Social Media. Contextualized versus Structural Overlapping Communities in Social Media.
Contextualized versus Structural Overlapping Communities in Social Media.
 
DeLiddo&BuckinghamShum-e-Part2014
DeLiddo&BuckinghamShum-e-Part2014DeLiddo&BuckinghamShum-e-Part2014
DeLiddo&BuckinghamShum-e-Part2014
 
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...
Comtella: Adaptive Rewards Mechanism to Incentivize Participation in Online C...
 
Centrality in Time- Dependent Networks
Centrality in Time- Dependent NetworksCentrality in Time- Dependent Networks
Centrality in Time- Dependent Networks
 
Paper Writing in Applied Mathematics (slightly updated slides)
Paper Writing in Applied Mathematics (slightly updated slides)Paper Writing in Applied Mathematics (slightly updated slides)
Paper Writing in Applied Mathematics (slightly updated slides)
 
Social Network Analysis (Part 1)
Social Network Analysis (Part 1)Social Network Analysis (Part 1)
Social Network Analysis (Part 1)
 
The Mathematics of Social Network Analysis: Metrics for Academic Social Networks
The Mathematics of Social Network Analysis: Metrics for Academic Social NetworksThe Mathematics of Social Network Analysis: Metrics for Academic Social Networks
The Mathematics of Social Network Analysis: Metrics for Academic Social Networks
 
4C13 J.15 Larson "Twitter based discourse community"
4C13 J.15 Larson "Twitter based discourse community"4C13 J.15 Larson "Twitter based discourse community"
4C13 J.15 Larson "Twitter based discourse community"
 
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...
Hypertext2017-Leveraging Followee List Memberships for Inferring User Interes...
 

Andere mochten auch

'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote
'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote
'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynotekjanowicz
 
Práctica 4. Portafolio de Presentación.
Práctica 4. Portafolio de Presentación.Práctica 4. Portafolio de Presentación.
Práctica 4. Portafolio de Presentación.Rosa Cabedo
 
Relationship among Economic Growth, Internet Usage and Publication Productivi...
Relationship among Economic Growth, Internet Usage and Publication Productivi...Relationship among Economic Growth, Internet Usage and Publication Productivi...
Relationship among Economic Growth, Internet Usage and Publication Productivi...University of Malaya
 
Strategic Planning for Information Systems
Strategic Planning for Information SystemsStrategic Planning for Information Systems
Strategic Planning for Information SystemsRon_Janzen
 
Rogue Element Mini-Portfolio
Rogue Element Mini-PortfolioRogue Element Mini-Portfolio
Rogue Element Mini-Portfoliorogueelement
 
Eurozone crisis and policy failure (Cádiz)
Eurozone crisis and policy failure (Cádiz)Eurozone crisis and policy failure (Cádiz)
Eurozone crisis and policy failure (Cádiz)Luis López-Molina
 
Daily Stock Report Monday Morning, July 9, 2012
Daily Stock Report Monday Morning, July 9, 2012Daily Stock Report Monday Morning, July 9, 2012
Daily Stock Report Monday Morning, July 9, 2012James Stewart
 
Cynthia shanmugalingam on social innovation in india
Cynthia shanmugalingam on social innovation in indiaCynthia shanmugalingam on social innovation in india
Cynthia shanmugalingam on social innovation in indiaSocial Innovation Exchange
 
Union Budget 2012-2013 of India
Union Budget 2012-2013 of IndiaUnion Budget 2012-2013 of India
Union Budget 2012-2013 of IndiaChaahat Khattar
 
Corporate Governance: Board Effectiveness Framework for Directors
Corporate Governance: Board Effectiveness Framework for DirectorsCorporate Governance: Board Effectiveness Framework for Directors
Corporate Governance: Board Effectiveness Framework for DirectorsJim Crocker
 
Banglai candlesticks pattern er sohoj path
Banglai candlesticks pattern er sohoj pathBanglai candlesticks pattern er sohoj path
Banglai candlesticks pattern er sohoj pathDrMohammed Nizam Uddin
 
gold monetization
gold monetizationgold monetization
gold monetizationArpit Daga
 
World: Titanium - Market Report. Analysis And Forecast To 2025
World: Titanium - Market Report. Analysis And Forecast To 2025World: Titanium - Market Report. Analysis And Forecast To 2025
World: Titanium - Market Report. Analysis And Forecast To 2025IndexBox Marketing
 
Final music magazine draft
Final music magazine draftFinal music magazine draft
Final music magazine draftmarinmou07
 

Andere mochten auch (20)

'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote
'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote
'The Why, What, and How of Geo-Information Observatories' GeoRich2014 Keynote
 
Práctica 4. Portafolio de Presentación.
Práctica 4. Portafolio de Presentación.Práctica 4. Portafolio de Presentación.
Práctica 4. Portafolio de Presentación.
 
Hybrid Distatis
Hybrid DistatisHybrid Distatis
Hybrid Distatis
 
Relationship among Economic Growth, Internet Usage and Publication Productivi...
Relationship among Economic Growth, Internet Usage and Publication Productivi...Relationship among Economic Growth, Internet Usage and Publication Productivi...
Relationship among Economic Growth, Internet Usage and Publication Productivi...
 
Strategic Planning for Information Systems
Strategic Planning for Information SystemsStrategic Planning for Information Systems
Strategic Planning for Information Systems
 
2008 crisis
2008 crisis2008 crisis
2008 crisis
 
Phr. efruzhu 34 99 new theori̇es and laws 2
Phr.  efruzhu  34 99 new theori̇es  and  laws  2Phr.  efruzhu  34 99 new theori̇es  and  laws  2
Phr. efruzhu 34 99 new theori̇es and laws 2
 
Pragati Bhotika - CV
Pragati Bhotika - CVPragati Bhotika - CV
Pragati Bhotika - CV
 
Rogue Element Mini-Portfolio
Rogue Element Mini-PortfolioRogue Element Mini-Portfolio
Rogue Element Mini-Portfolio
 
Eurozone crisis and policy failure (Cádiz)
Eurozone crisis and policy failure (Cádiz)Eurozone crisis and policy failure (Cádiz)
Eurozone crisis and policy failure (Cádiz)
 
Recognition From Top Brass Of Indus Ind
Recognition From Top Brass Of Indus IndRecognition From Top Brass Of Indus Ind
Recognition From Top Brass Of Indus Ind
 
Daily Stock Report Monday Morning, July 9, 2012
Daily Stock Report Monday Morning, July 9, 2012Daily Stock Report Monday Morning, July 9, 2012
Daily Stock Report Monday Morning, July 9, 2012
 
Cynthia shanmugalingam on social innovation in india
Cynthia shanmugalingam on social innovation in indiaCynthia shanmugalingam on social innovation in india
Cynthia shanmugalingam on social innovation in india
 
European Debt Crisis
European Debt CrisisEuropean Debt Crisis
European Debt Crisis
 
Union Budget 2012-2013 of India
Union Budget 2012-2013 of IndiaUnion Budget 2012-2013 of India
Union Budget 2012-2013 of India
 
Corporate Governance: Board Effectiveness Framework for Directors
Corporate Governance: Board Effectiveness Framework for DirectorsCorporate Governance: Board Effectiveness Framework for Directors
Corporate Governance: Board Effectiveness Framework for Directors
 
Banglai candlesticks pattern er sohoj path
Banglai candlesticks pattern er sohoj pathBanglai candlesticks pattern er sohoj path
Banglai candlesticks pattern er sohoj path
 
gold monetization
gold monetizationgold monetization
gold monetization
 
World: Titanium - Market Report. Analysis And Forecast To 2025
World: Titanium - Market Report. Analysis And Forecast To 2025World: Titanium - Market Report. Analysis And Forecast To 2025
World: Titanium - Market Report. Analysis And Forecast To 2025
 
Final music magazine draft
Final music magazine draftFinal music magazine draft
Final music magazine draft
 

Ähnlich wie Exploring Generative Models of Tripartite Graphs for Recommendation in Social Media

Beyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesBeyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesMaya Hristakeva
 
CORE Analytics Dashboard
CORE Analytics DashboardCORE Analytics Dashboard
CORE Analytics Dashboardpetrknoth
 
The state of the art in integrating machine learning into visual analytics
The state of the art in integrating machine learning into visual analyticsThe state of the art in integrating machine learning into visual analytics
The state of the art in integrating machine learning into visual analyticsCagatay Turkay
 
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...GUANGYUAN PIAO
 
Activating Research Collaboratories with Collaboration Patterns
Activating Research Collaboratories with Collaboration PatternsActivating Research Collaboratories with Collaboration Patterns
Activating Research Collaboratories with Collaboration PatternsCommunitySense
 
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...GUANGYUAN PIAO
 
Data Mining In Social Networks Using K-Means Clustering Algorithm
Data Mining In Social Networks Using K-Means Clustering AlgorithmData Mining In Social Networks Using K-Means Clustering Algorithm
Data Mining In Social Networks Using K-Means Clustering Algorithmnishant24894
 
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...Dominik Kowald
 
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTAL
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTALCONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTAL
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTALcscpconf
 
Contextual model of recommending resources on an academic networking portal
Contextual model of recommending resources on an academic networking portalContextual model of recommending resources on an academic networking portal
Contextual model of recommending resources on an academic networking portalcsandit
 
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...Lippo Group Digital
 
bookrecommendations-230615063942-3b1016c9 (1).pdf
bookrecommendations-230615063942-3b1016c9 (1).pdfbookrecommendations-230615063942-3b1016c9 (1).pdf
bookrecommendations-230615063942-3b1016c9 (1).pdf13DikshaDatir
 
IOTA 2016 Social Recomender System Presentation.
IOTA 2016 Social Recomender System Presentation.IOTA 2016 Social Recomender System Presentation.
IOTA 2016 Social Recomender System Presentation.ASHISH JAGTAP
 
Domain Modeling for Personalized Learning
Domain Modeling for Personalized LearningDomain Modeling for Personalized Learning
Domain Modeling for Personalized LearningPeter Brusilovsky
 
Predicting Communication Intention in Social Media
Predicting Communication Intention in Social MediaPredicting Communication Intention in Social Media
Predicting Communication Intention in Social MediaCharalampos Chelmis
 
Cold-Start Management with Cross-Domain Collaborative Filtering and Tags
Cold-Start Management with Cross-Domain Collaborative Filtering and TagsCold-Start Management with Cross-Domain Collaborative Filtering and Tags
Cold-Start Management with Cross-Domain Collaborative Filtering and TagsMatthias Braunhofer
 

Ähnlich wie Exploring Generative Models of Tripartite Graphs for Recommendation in Social Media (20)

Beyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research ArticlesBeyond Collaborative Filtering: Learning to Rank Research Articles
Beyond Collaborative Filtering: Learning to Rank Research Articles
 
CORE Analytics Dashboard
CORE Analytics DashboardCORE Analytics Dashboard
CORE Analytics Dashboard
 
The state of the art in integrating machine learning into visual analytics
The state of the art in integrating machine learning into visual analyticsThe state of the art in integrating machine learning into visual analytics
The state of the art in integrating machine learning into visual analytics
 
Slides ecir2016
Slides ecir2016Slides ecir2016
Slides ecir2016
 
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...
SEMANTiCS2016 - Exploring Dynamics and Semantics of User Interests for User ...
 
Activating Research Collaboratories with Collaboration Patterns
Activating Research Collaboratories with Collaboration PatternsActivating Research Collaboratories with Collaboration Patterns
Activating Research Collaboratories with Collaboration Patterns
 
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...
EKAW2016 - Interest Representation, Enrichment, Dynamics, and Propagation: A ...
 
Data Mining In Social Networks Using K-Means Clustering Algorithm
Data Mining In Social Networks Using K-Means Clustering AlgorithmData Mining In Social Networks Using K-Means Clustering Algorithm
Data Mining In Social Networks Using K-Means Clustering Algorithm
 
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...
WWW'15: A Hybrid Resource Recommender Mimicking Attention-Interpretation Dyna...
 
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTAL
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTALCONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTAL
CONTEXTUAL MODEL OF RECOMMENDING RESOURCES ON AN ACADEMIC NETWORKING PORTAL
 
Contextual model of recommending resources on an academic networking portal
Contextual model of recommending resources on an academic networking portalContextual model of recommending resources on an academic networking portal
Contextual model of recommending resources on an academic networking portal
 
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...
Profiler for Smartphone Users Interests Using Modified Hierarchical Agglomera...
 
bookrecommendations-230615063942-3b1016c9 (1).pdf
bookrecommendations-230615063942-3b1016c9 (1).pdfbookrecommendations-230615063942-3b1016c9 (1).pdf
bookrecommendations-230615063942-3b1016c9 (1).pdf
 
Book Recommendations.pptx
Book Recommendations.pptxBook Recommendations.pptx
Book Recommendations.pptx
 
Benoit Visual Only Retrieval
Benoit Visual Only RetrievalBenoit Visual Only Retrieval
Benoit Visual Only Retrieval
 
IOTA 2016 Social Recomender System Presentation.
IOTA 2016 Social Recomender System Presentation.IOTA 2016 Social Recomender System Presentation.
IOTA 2016 Social Recomender System Presentation.
 
Domain Modeling for Personalized Learning
Domain Modeling for Personalized LearningDomain Modeling for Personalized Learning
Domain Modeling for Personalized Learning
 
WWW2015 PHD Symposium
WWW2015 PHD SymposiumWWW2015 PHD Symposium
WWW2015 PHD Symposium
 
Predicting Communication Intention in Social Media
Predicting Communication Intention in Social MediaPredicting Communication Intention in Social Media
Predicting Communication Intention in Social Media
 
Cold-Start Management with Cross-Domain Collaborative Filtering and Tags
Cold-Start Management with Cross-Domain Collaborative Filtering and TagsCold-Start Management with Cross-Domain Collaborative Filtering and Tags
Cold-Start Management with Cross-Domain Collaborative Filtering and Tags
 

Kürzlich hochgeladen

Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024BookNet Canada
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
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
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machinePadma Pradeep
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsMemoori
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsMark Billinghurst
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticscarlostorres15106
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking MenDelhi Call girls
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Alan Dix
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptxLBM Solutions
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAndikSusilo4
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesSinan KOZAK
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdfhans926745
 

Kürzlich hochgeladen (20)

Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
 
A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
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
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machine
 
AI as an Interface for Commercial Buildings
AI as an Interface for Commercial BuildingsAI as an Interface for Commercial Buildings
AI as an Interface for Commercial Buildings
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Human Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR SystemsHuman Factors of XR: Using Human Factors to Design XR Systems
Human Factors of XR: Using Human Factors to Design XR Systems
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmaticsKotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
Kotlin Multiplatform & Compose Multiplatform - Starter kit for pragmatics
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
 
Key Features Of Token Development (1).pptx
Key  Features Of Token  Development (1).pptxKey  Features Of Token  Development (1).pptx
Key Features Of Token Development (1).pptx
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & Application
 
Unblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen FramesUnblocking The Main Thread Solving ANRs and Frozen Frames
Unblocking The Main Thread Solving ANRs and Frozen Frames
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 

Exploring Generative Models of Tripartite Graphs for Recommendation in Social Media

  • 1. Exploring Generative Models of Tripartite Graphs for Recommendation in Social Media Charalampos Chelmis, Viktor K Prasanna chelmis@usc.edu MSM 2013, Paris, France
  • 2. • Introduction • Structure of Tripartite Graphs • Generative Models of Tripartite Graphs • Social Link Classification Schemes • Evaluation • Conclusion Overview 2
  • 3. • Social Networking is used for  Content organization  Content sharing • Multiple media types • Users' activities  Reveal interests and tastes  Hidden structure • Description of Resources  Text  Tags / Hashtags • Social Annotation  Collective characterization of resources  Use of synonyms for similar recourses  Same keywords for different recourses Introduction 3
  • 4. • How to address issues of synonymy and polysemy?  Deal with space size explosion • How to discover emergent structure in online tagging systems?  Hidden topics • How to capture users’ latent interests?  Which subjects a user is mostly interested in?  Which users have similar interests? • How to model the process of social generation of annotations?  How to capture the semantics of collaboration • Why is this useful?  Recommend people  Recommend Tags / resources  Clustering  … Research Questions 4
  • 5. • Set of actors (e.g. users) A={a1, ...,ak} • Set of concepts (e.g. tags) C = {c1, ..., cl} • Set of resources (e.g. photos) R ={r1, ..., rm} Structure of Tripartite Graphs 5
  • 6. • The User-Concept Model  Users are modeled based on their tag usage  φ denotes the matrix of topic distributions − multinomial distribution over N concepts − T topics being drawn independently  θ: the matrix of user-specific mixture weights for these T topics • Captures users’ latent interests • Ignores Resources • Ignores the social aspect of tagging • The User-Resource Model  Resources become vocabulary terms • Tags are ignored • Ignores the social aspect of tagging Reducing the Tripartite Graph to Bipartite Structures 6
  • 7. • Topic-based representation • Model both resources & users’ interests • Multiple users may annotate resource r • For each tag a user is chosen uniformly at random • Each user is associated with a distribution over latent topics ɵ • A topic is chosen from a distribution over topics specific to that user • The tag is generated from the chosen topic  φt: probability distribution of tags for topic t The User-Resource-Concept Model 7
  • 8. • Tag Recommendation  Automatic annotation enhancement  Search improvement • Clustering  Community detection  Organization of resources/tags in categories • Navigation and Visualization  Social browsing • Next we focus on recommending people Recommendation 8
  • 9. • Classification Based on Latent Interests  Measure “tastes” distance with respect to latent topics distribution  Pointwise squared distance between feature vectors of users u and v   Other measures to consider − Kullback Leibler (KL) divergence − Cosine similarity • Objective:  Minimize the distance between linked users • Focus on topical homophily  Ignore network effects • Prior work uses network proximity as indicator of link formation Social Link Recommendation Using Latent Semantics & Network Structure 9 ]v))(k,-u)(k,(,,v))(1,-u)(1,[(v)F(u, 22 ΘΘΘΘ=  F(u,v) = 0 => u,v have identical distributions F(u,v) > 0 => distributions diverge
  • 10. • Latent Topics & Local Structure  CN(u,v) = common neighbors between users u and v − Simplicity and computational efficiency  Latent topics similarity   • Latent Topics & Global Structure  SD(u,v) = shortest distance between users u and v  • Non separable training set => inefficient classifiers • Aggregation Strategy  Reduce the number of training samples  Produce more efficient classifiers  Average latent similarity of user pairs with k common neighbors: Social Link Recommendation Using Latent Semantics & Network Structure 10 v)]CN(u,v),(u,[v)F(u, σ= ∑== = kk:pp p (p) |kk:p| 1 (k)avg σσ v)]SD(u,v),(u,[v)F(u, σ= 22 ),(),( ),(),( ),( ∑∑ ∑ ΘΘ ΘΘ = tt t vtut vtut vuσ
  • 11. • Objectives  Ability to uncover subliminal collective knowledge  Evaluate performance of “people” recommendation • Setting  2.4 GHz Intel Core 2 Duo, 2 GB memory, Windows 7 • Real-world Dataset  Last.fm online music system − social relationships − tagging information − music artist listening information  Statistics − 1,892 users − 25,434 directed user friend relations − 17,632 artists UR Model vocabulary size − 92,834 user-listened-artist relations − 11,946 unique tags UC and URC vocabulary size − 186,479 annotations (tuples <user, tag, artist>) Experimental Analysis 11
  • 13. • Evaluate ability to predict tags/resources on new users  Perplexity • Split dataset into two disjoint sets  90% for training • Lower perplexity indicates better generalization • URC better overall  Exploits more information • UC  Organizes tags in “clusters” • UR  Inferior quality due to noise Predictive Power 13
  • 14. • Split dataset into two disjoint sets  10%, 25%, 50%, 75% for training, rest for testing • Evaluation process  Randomly sample 12,716 pairs of users  50% true links, 50% negative samples  Compute similarity of user pairs  Sort users in decreasing order of similarity  Add links between users with highest similarity Recommendation of Social Ties 14
  • 15. • Latent Topics & Shortest Distance  Aggregates all true links training similarity values in a single point  Least effective • Ensemble achieves best precision • Over fitting for training size > 50% • Recall drops as dataset size increases Recommendation of Social Ties 15 [Latent Topics & Local Structure] [Latent Topics] [Ensemble]
  • 16. • In social media number of true links << absent links • High performance for both classes  True negatives easier to classify correctly  Degradation in performance for true positives • Reasonable results for practical purposes How about High Class Imbalance? 16 [Latent Topics & Local Structure] [Latent Topics] [Ensemble]
  • 17. • Baselines  Cosine Similarity (CS)  Maximal Information Path (MIP) • Evaluation Criterion  Area under the receiver-operating characteristic curve (AUC) • Baselines AUC  Computed over the complete dataset  Biases the evaluation in favor of the baselines  CS AUC = 0.6087  MIP AUC = 0.6256 • Same evaluation process as before • Compute performance lift  % change over best performing baseline  Positive % denotes improvement Comparison to Tag-based similarity metrics 17
  • 18. • Not all schemes can beat the baseline  For 10% training data  ≤10% AUC loss  But, significant speedup due to minimal training dataset • Latent Topics & Local Structure Scheme consistently better Comparison to Tag-based similarity metrics 18 Training dataset size [Latent Topics & Local Structure] [Latent Topics]
  • 19. • Three generative models of tripartite graphs in social tagging systems • Modeling of users’ interests in a latent space over resources and metadata • Limitations  Ignore several aspects of real-world annotation process, such as topic correlation and user interaction • Achieve great performance in the recommendation task  Accurate predictors of social ties in conjunction with structural evidence  Proposed aggregation strategy to reduce number of training samples • Future work  Incorporate other types of resources  Automatically identify most discriminative latent topics and discard uninformative resources and metadata Concluding Remarks 19