SlideShare a Scribd company logo
1 of 38
Download to read offline
The Role of Emotions in Context-aware Recommendation
Yong Zheng, Robin Burke, Bamshad Mobasher
Center for Web Intelligence, DePaul University, USA
Decisions@RecSys 2013, Hong Kong, Oct 12
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
Recommender Systems (RS)
• Information overload problem
• Adapt user preference
• Provide a list of recommendations
• Assist decision makings
E-Commerce
(Amazon.com)
Social RS (Twitter) Tagging RS (Flickr)
Examples in other domains: Netflix, Pandora, Yelp, etc
Affective Recommender Systems
• Affective computing is the study and development
of systems and devices that can recognize,
interpret, process, and simulate human affects. It is
an interdisciplinary field spanning computer
sciences, psychology, and cognitive science.
• Decision Making = Rational + Emotional
(Daniel Kahnemann, Economic Nobel Prize 2002)
• Affective Recommender System takes emotional
effects into consideration to further assist
decision makings and recommendations.
E.g. music preferences by emotional effect
M. Tkalcic, A. Kosir, J. Tasic. "Affective recommender systems: the role of
emotions in recommender systems." Decisions@RecSys 2011
Take Emotions into RS
• Content-based Frameworks
Semantic emotions can be extracted from tags
(affective tags), reviews (review or opinion mining),
user tweets, metadata, etc
• Collaborative Filtering
In a way of “collaborative”: see how other users
prefer with similar emotions; or see how other
items user may prefer within similar emotions
• Context-aware Recommender Systems
It is a multidimensional approach, where
emotions are deemed as additional dimensions
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
Context-aware RS (CARS)
• Traditional RS: Users × Items  Ratings
• CARS: Users × Items × Contexts  Ratings
Context-aware RS (CARS)
• Example: Tour Plan Recommender
Yu, Chien-Chih and Chang, Hsiao-ping, "Towards Context-Aware Recommendation for
Personalized Mobile Travel Planning". International Conference on Context-Aware Systems
and Applications, ICCASA 2012
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
The Role of Emotions in CARS
• Several research has considered emotions as
contexts and demonstrate the improved
efficiency in contextual recommendation
• But what is the role of emotions?
Improved effectiveness in terms of MAE, RMSE,
Precision, etc cannot fully reveal the roles.
The Role of Emotions in CARS
Definition: the role of emotions
• Effectiveness: how effective the emotions are?
Comparison among emotion only, no
emotions, and a mixture of emotions &
other features.
• Usage: how emotions are used in CARS? E.g.
which emotions are selected?
which emotions are most influential ones?
which components emotions are applied to?
any latent patterns in terms of usage?
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
Explore The Role of Emotions in CARS
• Effectiveness: how effective the emotions are?
Comparison among emotion only, no
emotions, and a mixture of emotions &
other features.
• Usage: how emotions are used in CARS? E.g.
which emotions are selected?
which emotions are most influential ones?
which components emotions are applied to?
any latent patterns in terms of usage?
• How to explore?
Two classes of CARS algorithms:
Context-aware Splitting Approaches;
Differential Context Modeling;
Baseline: Context-aware Matrix Factorization
Context-aware Splitting Approaches
• Three Algorithms:
Item Splitting, User Splitting, UI Splitting
• Item Splitting: The underlying idea is that the
nature of an item, from the user's point of view,
may change in different contextual conditions,
hence it may be useful to consider it as two
different items. (L. Baltrunas, F. Ricci, RecSys'09)
At Cinema At Home At Swimming Pool
Context-aware Splitting Approaches
Splitting Process:
Choose a contextual condition to split each item;
The selection process is done by measuring significance
of rating differences (such as the two-sample t test);
Then multidimensional matrix is converted to a 2D one,
then traditional algorithms like CF, MF can be applied to;
User Item Loc Rating
U1 M1 Pool 5
U2 M1 Pool 5
U3 M1 Pool 5
U1 M1 Home 2
U4 M1 Home 3
U2 M1 Cinema 2
High Ratings
Low Ratings
Significant difference?
Let’s split it !!!
Context-aware Splitting Approaches
Matrix Transformation
User Item Loc Rating
U1 M1 Pool 5
U2 M1 Pool 5
U3 M1 Pool 5
U1 M1 Home 2
U4 M1 Home 3
U2 M1 Cinema 2
User Item Rating
U1 M11 5
U2 M11 5
U3 M11 5
U1 M12 2
U4 M12 3
U2 M12 2
Transformation
An alternative
contextual condition
for splitting:
“Pool” vs. “Non-Pool”
If there is qualified
split, one item will
be split to two new
ones.
 Contexts are fused into items as new items in 2D rating matrix;
 Traditional RS algorithms can be applied to;
Context-aware Splitting Approaches
• User Splitting: is a similar one. Instead of splitting
items, it may be useful to consider one user as two
different users, if user demonstrates significantly
different preferences across contexts. (L. Baltrunas,
et al., CARS@RecSys 2009 and A. Said et al.,
CARS@RecSys 2011)
• UI Splitting: simply a combination of item splitting
and user splitting – both approaches are applied to
create a new rating matrix – new users and new
items are created in the rating matrix. (Y. Zheng, R.
Burke, B. Mobasher, Decisions@RecSys 2013)
Context-aware Splitting Approaches
Context-aware Splitting Approaches
How about the characteristics of this approach in terms of discovering
the usage of contexts in the recommendation process?
Context-aware splitting approaches have to select influential contextual
conditions to split users or items; so the usage can be inferred by the
distribution of which emotional contexts are used to split items and users,
and which ones are the most frequent used ones?
Differential Context Modeling
 Separate one algorithm to different components;
 Apply optimal contextual constraints to each component. The underlying is that, one
context may be influential for this component, but it is not usually the case for other
components.
Differential Context Modeling
Contextual constraints can be applied by a condition-matching (differential context
relaxation, DCR) or a similarity-matching (differential context weighting, DCW); In DCR, the
contexts should exactly matched by selected variables. In DCW, a similar enough context is
allowed – it is not necessary to be exactly matched, but should be similar enough.
Location: Home, Cinema, Swimming Pool
Time: Weekend, Weekday
Companion: Family, Friend, Girlfriend
If location and time are selected:
DCR  exactly match {Cinema, Weekend}
DCW{Home, Weekend} is 50% similar!
(Simply, assume dimensions have equal weights in DCW)
Differential Context Modeling
How about the characteristics of this approach in terms of discovering
the usage of contexts in the recommendation process?
DCR  Which emotional dimensions are selected for each component? (feature selection)
DCW  Which dimensions are weighted higher in each component? (feature weighting)
Baseline in Performance: CAMF
CAMF = Context-aware Matrix Factorization
We just applied CAMF to compare predictive performance, and did
not explore the usage of contexts in this approach.
Basic MF:
Bias MF:
CAMF_CI:
CAMF_CU:
Reference: Baltrunas, Linas, Bernd Ludwig, and Francesco Ricci. "Matrix
factorization techniques for context aware recommendation." Proceedings of
the fifth ACM conference on Recommender systems. ACM, 2011.
Three Shapes: CAMF_C, CAMF_CI, CAMF_CU
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
Experimental Design
Contextual variables in the Dataset: LDOS-CoMoDa
Emotional contexts
# of users: 113; # of items: 1186; # of ratings: 2094; rating scale: 1-5
Each user has at least 5 ratings profiles in the data;
Experimental Design
• Effectiveness: how effective the emotions are?
Comparison among emotion only, no
emotions, and a mixture of emotions &
other features (all contexts).
• Usage: how emotions are used in CARS? E.g.
which emotions are selected?
which emotions are most influential ones?
which components emotions are applied to?
any latent patterns in terms of usage?
• Effectiveness: Comparison among emotion only, no emotions, and a
mixture of emotions & other features (all contexts) by RMSE.
Findings:
 “No emotions” plays the worst; (Note: MF is applied to splitting approaches)
 “Emotion Only” works the best for CAMF; “All contexts” is the best for other two approaches;
 UI splitting achieves lowest RMSE in all situations: All contexts, no emotions & emotion only;
 User splitting performs better than item splitting for this data;
 Emotions are beneficial !!!
Experimental Design
• Effectiveness: how effective the emotions are?
Comparison among emotion only, no
emotions, and a mixture of emotions &
other features (all contexts).
• Usage: how emotions are used in CARS? E.g.
which emotions are selected?
which emotions are most influential ones?
which components emotions are applied to?
any latent patterns in terms of usage?
 Context-aware splitting approaches
the usage distribution of contexts used to split items or users
 Differential context modeling
DCR  Which dimensions are selected for each component?
DCW  Which dimensions are weighted higher in each component?
• Usage by Context-aware Splitting Approaches
Top-3 selections in item splitting: EndEmotion, Time, Season
Top-3 selections in user splitting: EndEmotion, DominantEmo, Time
Note: User splitting outperforms item splitting, any patterns?
Conjecture: dependences between contexts and users|items.
Experimental Results
• Usage by Differential Context Modeling
Notice: we applied DCM to user-based Collaborative Filtering.
 DCR  Which contexts are selected for each component?
 DCW  Which contexts are weighted higher in each component?
For DCW, we did not show weights, but give selected dimensions weighted higher than 0.7.
In DCR, emotional variables are selected in components of user baseline and similarity;
In DCW, emotions are weighted significantly higher expect for neighbor contribution;
DCM is optimal algorithm in context selection and weighting. Once emotions are selected or
weighted highly, they are optimal solutions!
• Summary of the Usage Patterns
In Context-aware Splitting Approaches:
 EndEmo and DominantEmo are two most influential ones!
 And emotion variables defeat other variables! (topper selection!)
 User splitting works better than Item splitting for this data
In Differential Context Modeling:
 EndEmo and DominantEmo are selected in DCR, and all three
emotions are selected in DCW.
 EndEmo and DominantEmo are influential for two components in
both DCR and DCW: user baseline and user similarity calculation
Overlapped findings: EndEmo and DominantEmo are influential ones!
• Recommender Systems (RS) & Affective RS
• Context-aware Recommender Systems (CARS)
• The Role of Emotions in CARS
• Explore The Roles by CARS Algorithms
• Experimental Results
• Conclusions and Future Work
• What we did?
Emotions are demonstrated to improve recommendation performance, but few research focus on the
other side of the roles of emotions – the usage instead of performance only.
In this work, we explore the roles (effectiveness + usage) by two classes of CARS algorithm: context-
aware splitting approach, and differential context modeling
• What we found?
Effectiveness:
UI splitting performs the best among all CARS algorithm for this data set;
A mixture of all contexts is better for splitting approaches and DCM;
CAMF may require context selection in advance;
Usage:
Splitting methods can explore usage by distribution of how frequent one context is used
to split items or users; DCM infers the selection or the degree of importance by weights
for each algorithm component; It can tell the importance of contexts by algorithm components.
Patterns:
Among three emotional variables, DominantEmo & EndEmo are two influential ones;
Conjecture: User splitting may outperform item splitting if contexts are more dependent with
users than items; experiments on more data are required to confirm this pattern.
Future Work
 The emotions in this data are not real emotions, e.g. DominantEmo and EndEmo
are post-emotions, how about pre-emotions (emotion before the activity)?
 What are the insight of those emotions? Any associations among emotions, user
profiles and item features? For example, EndEmo is sad after seeing a tragedy, but
rating is high (because tragedy makes people cry…), Or, a happy EndEmo will
always results in higher rating? (because users likes this experience).
 In terms of the three splitting approaches, especially the new proposed one – UI
Splitting, it is useful to perform an empirical study among those three splitting
approaches to further explore in which situation, which one will perform better
than other ones.
References
Item Splitting
1) L. Baltrunas, and F. Ricci. "Context-based splitting of item ratings in collaborative filtering." RecSys, 2009.
2) L. Baltrunas, and F. Ricci. "Experimental evaluation of context-dependent collaborative filtering using item splitting." User
Modeling and User-Adapted Interaction (2013): 1-28.
User Splitting
1) L. Batrunas and X. Amatriain."Towards Time-Dependent Recommendation Based on Implicit Feedback." CARS@RecSys, 2009
2) A. Said, E. Luca, S. Albayrak. "Inferring contextual user profiles—improving recommender performance.“ CARS@RecSys, 2011
UI Splitting
1) Y. Zheng, R. Burke, B. Mobasher. "The Role of Emotions in Context-aware Recommendation". Decisons@RecSys, 2013
Context-aware Matrix Factorization
1) L. Baltrunas, B. Ludwig, F. Ricci. "Matrix factorization techniques for context aware recommendation." RecSys 2011.
Differential Context Modeling
1) Y. Zheng, R. Burke, B. Mobasher. "Differential Context Relaxation for Context-aware Travel Recommendation". EC-WEB, 2012
2) Y. Zheng, R. Burke, B. Mobasher. "Optimal Feature Selection for Context-Aware Recommendation using Differential Relaxation".
CARS@RecSys, 2012
3) Y. Zheng, R. Burke, B. Mobasher. "Recommendation with Differential Context Weighting". In UMAP, 2013
Thank You!
Decisions@RecSys 2013, Hong Kong
Center for Web Intelligence, DePaul University, USA
10/12/2013

More Related Content

What's hot

[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation ApproachYONG ZHENG
 
Tutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender SystemsTutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender SystemsYONG ZHENG
 
Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsTutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsYONG ZHENG
 
Movie lens movie recommendation system
Movie lens movie recommendation systemMovie lens movie recommendation system
Movie lens movie recommendation systemGaurav Sawant
 
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
[WI 2017] Context Suggestion: Empirical Evaluations vs User StudiesYONG ZHENG
 
Summary of a Recommender Systems Survey paper
Summary of a Recommender Systems Survey paperSummary of a Recommender Systems Survey paper
Summary of a Recommender Systems Survey paperChangsung Moon
 
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...YONG ZHENG
 
Matrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender SystemsMatrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender SystemsLei Guo
 
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...YONG ZHENG
 
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
[WI 2017] Affective Prediction By Collaborative Chains In Movie RecommendationYONG ZHENG
 
Latent factor models for Collaborative Filtering
Latent factor models for Collaborative FilteringLatent factor models for Collaborative Filtering
Latent factor models for Collaborative Filteringsscdotopen
 
(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project report(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project reportGaurav Sawant
 
Recommender Systems! @ASAI 2011
Recommender Systems! @ASAI 2011Recommender Systems! @ASAI 2011
Recommender Systems! @ASAI 2011Ernesto Mislej
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender SystemsLior Rokach
 
Stated preference methods and analysis
Stated preference methods and analysisStated preference methods and analysis
Stated preference methods and analysisHabet Madoyan
 
Movie recommendation project
Movie recommendation projectMovie recommendation project
Movie recommendation projectAbhishek Jaisingh
 
GTC 2021: Counterfactual Learning to Rank in E-commerce
GTC 2021: Counterfactual Learning to Rank in E-commerceGTC 2021: Counterfactual Learning to Rank in E-commerce
GTC 2021: Counterfactual Learning to Rank in E-commerceGrubhubTech
 
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...Alejandro Bellogin
 
Recommender system
Recommender systemRecommender system
Recommender systemSaiguru P.v
 
Movies recommendation system in R Studio, Machine learning
Movies recommendation system in  R Studio, Machine learning Movies recommendation system in  R Studio, Machine learning
Movies recommendation system in R Studio, Machine learning Mauryasuraj98
 

What's hot (20)

[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
[IUI 2017] Criteria Chains: A Novel Multi-Criteria Recommendation Approach
 
Tutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender SystemsTutorial: Context-awareness In Information Retrieval and Recommender Systems
Tutorial: Context-awareness In Information Retrieval and Recommender Systems
 
Tutorial: Context In Recommender Systems
Tutorial: Context In Recommender SystemsTutorial: Context In Recommender Systems
Tutorial: Context In Recommender Systems
 
Movie lens movie recommendation system
Movie lens movie recommendation systemMovie lens movie recommendation system
Movie lens movie recommendation system
 
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
[WI 2017] Context Suggestion: Empirical Evaluations vs User Studies
 
Summary of a Recommender Systems Survey paper
Summary of a Recommender Systems Survey paperSummary of a Recommender Systems Survey paper
Summary of a Recommender Systems Survey paper
 
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
[RIIT 2017] Identifying Grey Sheep Users By The Distribution of User Similari...
 
Matrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender SystemsMatrix Factorization Techniques For Recommender Systems
Matrix Factorization Techniques For Recommender Systems
 
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
[ADMA 2017] Identification of Grey Sheep Users By Histogram Intersection In R...
 
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
[WI 2017] Affective Prediction By Collaborative Chains In Movie Recommendation
 
Latent factor models for Collaborative Filtering
Latent factor models for Collaborative FilteringLatent factor models for Collaborative Filtering
Latent factor models for Collaborative Filtering
 
(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project report(Gaurav sawant & dhaval sawlani)bia 678 final project report
(Gaurav sawant & dhaval sawlani)bia 678 final project report
 
Recommender Systems! @ASAI 2011
Recommender Systems! @ASAI 2011Recommender Systems! @ASAI 2011
Recommender Systems! @ASAI 2011
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Stated preference methods and analysis
Stated preference methods and analysisStated preference methods and analysis
Stated preference methods and analysis
 
Movie recommendation project
Movie recommendation projectMovie recommendation project
Movie recommendation project
 
GTC 2021: Counterfactual Learning to Rank in E-commerce
GTC 2021: Counterfactual Learning to Rank in E-commerceGTC 2021: Counterfactual Learning to Rank in E-commerce
GTC 2021: Counterfactual Learning to Rank in E-commerce
 
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...
HT2014 Tutorial: Evaluating Recommender Systems - Ensuring Replicability of E...
 
Recommender system
Recommender systemRecommender system
Recommender system
 
Movies recommendation system in R Studio, Machine learning
Movies recommendation system in  R Studio, Machine learning Movies recommendation system in  R Studio, Machine learning
Movies recommendation system in R Studio, Machine learning
 

Similar to The Role of Emotions in Context-Aware Recommendation Systems

Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender systemStanley Wang
 
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...Kishor Datta Gupta
 
[WI 2014]Context Recommendation Using Multi-label Classification
[WI 2014]Context Recommendation Using Multi-label Classification[WI 2014]Context Recommendation Using Multi-label Classification
[WI 2014]Context Recommendation Using Multi-label ClassificationYONG ZHENG
 
Collaborative Filtering Recommendation System
Collaborative Filtering Recommendation SystemCollaborative Filtering Recommendation System
Collaborative Filtering Recommendation SystemMilind Gokhale
 
session2.pdf
session2.pdfsession2.pdf
session2.pdfshero2015
 
Advanced topics research
Advanced topics researchAdvanced topics research
Advanced topics researchkieran122
 
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNING
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNINGENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNING
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNINGIRJET Journal
 
Recommendation system (1).pptx
Recommendation system (1).pptxRecommendation system (1).pptx
Recommendation system (1).pptxprathammishra28
 
recommendationsystem1-221109055232-c8b46131.pdf
recommendationsystem1-221109055232-c8b46131.pdfrecommendationsystem1-221109055232-c8b46131.pdf
recommendationsystem1-221109055232-c8b46131.pdf13DikshaDatir
 
Teacher training material
Teacher training materialTeacher training material
Teacher training materialVikram Parmar
 
Lecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionLecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionPerumalPitchandi
 
FIND MY VENUE: Content & Review Based Location Recommendation System
FIND MY VENUE: Content & Review Based Location Recommendation SystemFIND MY VENUE: Content & Review Based Location Recommendation System
FIND MY VENUE: Content & Review Based Location Recommendation SystemIJTET Journal
 
IRJET- Personalize Travel Recommandation based on Facebook Data
IRJET- Personalize Travel Recommandation based on Facebook DataIRJET- Personalize Travel Recommandation based on Facebook Data
IRJET- Personalize Travel Recommandation based on Facebook DataIRJET Journal
 
How to use LLMs for creating a content-based recommendation system for entert...
How to use LLMs for creating a content-based recommendation system for entert...How to use LLMs for creating a content-based recommendation system for entert...
How to use LLMs for creating a content-based recommendation system for entert...mahaffeycheryld
 
Hybrid Deep Learning Model for Multilingual Sentiment Analysis
Hybrid Deep Learning Model for Multilingual Sentiment AnalysisHybrid Deep Learning Model for Multilingual Sentiment Analysis
Hybrid Deep Learning Model for Multilingual Sentiment AnalysisIRJET Journal
 
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...ijiert bestjournal
 
Dynamic personalized recommendation on sparse data
Dynamic personalized recommendation on sparse dataDynamic personalized recommendation on sparse data
Dynamic personalized recommendation on sparse dataJPINFOTECH JAYAPRAKASH
 

Similar to The Role of Emotions in Context-Aware Recommendation Systems (20)

Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender system
 
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
Deep Reinforcement Learning based Recommendation with Explicit User-ItemInter...
 
Abstractive Review Summarization
Abstractive Review SummarizationAbstractive Review Summarization
Abstractive Review Summarization
 
[WI 2014]Context Recommendation Using Multi-label Classification
[WI 2014]Context Recommendation Using Multi-label Classification[WI 2014]Context Recommendation Using Multi-label Classification
[WI 2014]Context Recommendation Using Multi-label Classification
 
Collaborative Filtering Recommendation System
Collaborative Filtering Recommendation SystemCollaborative Filtering Recommendation System
Collaborative Filtering Recommendation System
 
session2.pdf
session2.pdfsession2.pdf
session2.pdf
 
Advanced topics research
Advanced topics researchAdvanced topics research
Advanced topics research
 
PhD defense
PhD defense PhD defense
PhD defense
 
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNING
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNINGENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNING
ENTERTAINMENT CONTENT RECOMMENDATION SYSTEM USING MACHINE LEARNING
 
Recommendation system (1).pptx
Recommendation system (1).pptxRecommendation system (1).pptx
Recommendation system (1).pptx
 
recommendationsystem1-221109055232-c8b46131.pdf
recommendationsystem1-221109055232-c8b46131.pdfrecommendationsystem1-221109055232-c8b46131.pdf
recommendationsystem1-221109055232-c8b46131.pdf
 
NLP Ecosystem
NLP EcosystemNLP Ecosystem
NLP Ecosystem
 
Teacher training material
Teacher training materialTeacher training material
Teacher training material
 
Lecture Notes on Recommender System Introduction
Lecture Notes on Recommender System IntroductionLecture Notes on Recommender System Introduction
Lecture Notes on Recommender System Introduction
 
FIND MY VENUE: Content & Review Based Location Recommendation System
FIND MY VENUE: Content & Review Based Location Recommendation SystemFIND MY VENUE: Content & Review Based Location Recommendation System
FIND MY VENUE: Content & Review Based Location Recommendation System
 
IRJET- Personalize Travel Recommandation based on Facebook Data
IRJET- Personalize Travel Recommandation based on Facebook DataIRJET- Personalize Travel Recommandation based on Facebook Data
IRJET- Personalize Travel Recommandation based on Facebook Data
 
How to use LLMs for creating a content-based recommendation system for entert...
How to use LLMs for creating a content-based recommendation system for entert...How to use LLMs for creating a content-based recommendation system for entert...
How to use LLMs for creating a content-based recommendation system for entert...
 
Hybrid Deep Learning Model for Multilingual Sentiment Analysis
Hybrid Deep Learning Model for Multilingual Sentiment AnalysisHybrid Deep Learning Model for Multilingual Sentiment Analysis
Hybrid Deep Learning Model for Multilingual Sentiment Analysis
 
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...
A SURVEY PAPER ON EXTRACTION OF OPINION WORD AND OPINION TARGET FROM ONLINE R...
 
Dynamic personalized recommendation on sparse data
Dynamic personalized recommendation on sparse dataDynamic personalized recommendation on sparse data
Dynamic personalized recommendation on sparse data
 

More from YONG ZHENG

[WISE 2015] Similarity-Based Context-aware Recommendation
[WISE 2015] Similarity-Based Context-aware Recommendation[WISE 2015] Similarity-Based Context-aware Recommendation
[WISE 2015] Similarity-Based Context-aware RecommendationYONG ZHENG
 
Matrix Factorization In Recommender Systems
Matrix Factorization In Recommender SystemsMatrix Factorization In Recommender Systems
Matrix Factorization In Recommender SystemsYONG ZHENG
 
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...YONG ZHENG
 
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...YONG ZHENG
 
[UMAP2013] Recommendation with Differential Context Weighting
[UMAP2013] Recommendation with Differential Context Weighting[UMAP2013] Recommendation with Differential Context Weighting
[UMAP2013] Recommendation with Differential Context WeightingYONG ZHENG
 
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative FilteringYONG ZHENG
 
Slope one recommender on hadoop
Slope one recommender on hadoopSlope one recommender on hadoop
Slope one recommender on hadoopYONG ZHENG
 
A manual for Ph.D dissertation
A manual for Ph.D dissertationA manual for Ph.D dissertation
A manual for Ph.D dissertationYONG ZHENG
 
Attention flow by tagging prediction
Attention flow by tagging predictionAttention flow by tagging prediction
Attention flow by tagging predictionYONG ZHENG
 
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...YONG ZHENG
 
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...YONG ZHENG
 
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...YONG ZHENG
 

More from YONG ZHENG (12)

[WISE 2015] Similarity-Based Context-aware Recommendation
[WISE 2015] Similarity-Based Context-aware Recommendation[WISE 2015] Similarity-Based Context-aware Recommendation
[WISE 2015] Similarity-Based Context-aware Recommendation
 
Matrix Factorization In Recommender Systems
Matrix Factorization In Recommender SystemsMatrix Factorization In Recommender Systems
Matrix Factorization In Recommender Systems
 
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
[SAC2014]Splitting Approaches for Context-Aware Recommendation: An Empirical ...
 
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
[UMAP2013]Tutorial on Context-Aware User Modeling for Recommendation by Bamsh...
 
[UMAP2013] Recommendation with Differential Context Weighting
[UMAP2013] Recommendation with Differential Context Weighting[UMAP2013] Recommendation with Differential Context Weighting
[UMAP2013] Recommendation with Differential Context Weighting
 
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering[SOCRS2013]Differential Context Modeling in Collaborative Filtering
[SOCRS2013]Differential Context Modeling in Collaborative Filtering
 
Slope one recommender on hadoop
Slope one recommender on hadoopSlope one recommender on hadoop
Slope one recommender on hadoop
 
A manual for Ph.D dissertation
A manual for Ph.D dissertationA manual for Ph.D dissertation
A manual for Ph.D dissertation
 
Attention flow by tagging prediction
Attention flow by tagging predictionAttention flow by tagging prediction
Attention flow by tagging prediction
 
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
[CARS2012@RecSys]Optimal Feature Selection for Context-Aware Recommendation u...
 
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
[ECWEB2012]Differential Context Relaxation for Context-Aware Travel Recommend...
 
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
[HetRec2011@RecSys]Experience Discovery: Hybrid Recommendation of Student Act...
 

Recently uploaded

Take control of your SAP testing with UiPath Test Suite
Take control of your SAP testing with UiPath Test SuiteTake control of your SAP testing with UiPath Test Suite
Take control of your SAP testing with UiPath Test SuiteDianaGray10
 
The Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsThe Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsPixlogix Infotech
 
Long journey of Ruby standard library at RubyConf AU 2024
Long journey of Ruby standard library at RubyConf AU 2024Long journey of Ruby standard library at RubyConf AU 2024
Long journey of Ruby standard library at RubyConf AU 2024Hiroshi SHIBATA
 
Data governance with Unity Catalog Presentation
Data governance with Unity Catalog PresentationData governance with Unity Catalog Presentation
Data governance with Unity Catalog PresentationKnoldus Inc.
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfMounikaPolabathina
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPathCommunity
 
Potential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsPotential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsRavi Sanghani
 
Scale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterScale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterMydbops
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesKari Kakkonen
 
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxThe Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxLoriGlavin3
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxLoriGlavin3
 
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024BookNet Canada
 
TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024Lonnie McRorey
 
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data PrivacyTrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data PrivacyTrustArc
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .Alan Dix
 
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better StrongerModern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better Strongerpanagenda
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityIES VE
 
Generative Artificial Intelligence: How generative AI works.pdf
Generative Artificial Intelligence: How generative AI works.pdfGenerative Artificial Intelligence: How generative AI works.pdf
Generative Artificial Intelligence: How generative AI works.pdfIngrid Airi González
 
Sample pptx for embedding into website for demo
Sample pptx for embedding into website for demoSample pptx for embedding into website for demo
Sample pptx for embedding into website for demoHarshalMandlekar2
 
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality AssuranceInflectra
 

Recently uploaded (20)

Take control of your SAP testing with UiPath Test Suite
Take control of your SAP testing with UiPath Test SuiteTake control of your SAP testing with UiPath Test Suite
Take control of your SAP testing with UiPath Test Suite
 
The Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and ConsThe Ultimate Guide to Choosing WordPress Pros and Cons
The Ultimate Guide to Choosing WordPress Pros and Cons
 
Long journey of Ruby standard library at RubyConf AU 2024
Long journey of Ruby standard library at RubyConf AU 2024Long journey of Ruby standard library at RubyConf AU 2024
Long journey of Ruby standard library at RubyConf AU 2024
 
Data governance with Unity Catalog Presentation
Data governance with Unity Catalog PresentationData governance with Unity Catalog Presentation
Data governance with Unity Catalog Presentation
 
What is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdfWhat is DBT - The Ultimate Data Build Tool.pdf
What is DBT - The Ultimate Data Build Tool.pdf
 
UiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to HeroUiPath Community: Communication Mining from Zero to Hero
UiPath Community: Communication Mining from Zero to Hero
 
Potential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and InsightsPotential of AI (Generative AI) in Business: Learnings and Insights
Potential of AI (Generative AI) in Business: Learnings and Insights
 
Scale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL RouterScale your database traffic with Read & Write split using MySQL Router
Scale your database traffic with Read & Write split using MySQL Router
 
Testing tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examplesTesting tools and AI - ideas what to try with some tool examples
Testing tools and AI - ideas what to try with some tool examples
 
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptxThe Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
The Role of FIDO in a Cyber Secure Netherlands: FIDO Paris Seminar.pptx
 
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptxDigital Identity is Under Attack: FIDO Paris Seminar.pptx
Digital Identity is Under Attack: FIDO Paris Seminar.pptx
 
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
New from BookNet Canada for 2024: Loan Stars - Tech Forum 2024
 
TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024TeamStation AI System Report LATAM IT Salaries 2024
TeamStation AI System Report LATAM IT Salaries 2024
 
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data PrivacyTrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
 
From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .From Family Reminiscence to Scholarly Archive .
From Family Reminiscence to Scholarly Archive .
 
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better StrongerModern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
Modern Roaming for Notes and Nomad – Cheaper Faster Better Stronger
 
Decarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a realityDecarbonising Buildings: Making a net-zero built environment a reality
Decarbonising Buildings: Making a net-zero built environment a reality
 
Generative Artificial Intelligence: How generative AI works.pdf
Generative Artificial Intelligence: How generative AI works.pdfGenerative Artificial Intelligence: How generative AI works.pdf
Generative Artificial Intelligence: How generative AI works.pdf
 
Sample pptx for embedding into website for demo
Sample pptx for embedding into website for demoSample pptx for embedding into website for demo
Sample pptx for embedding into website for demo
 
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance[Webinar] SpiraTest - Setting New Standards in Quality Assurance
[Webinar] SpiraTest - Setting New Standards in Quality Assurance
 

The Role of Emotions in Context-Aware Recommendation Systems

  • 1. The Role of Emotions in Context-aware Recommendation Yong Zheng, Robin Burke, Bamshad Mobasher Center for Web Intelligence, DePaul University, USA Decisions@RecSys 2013, Hong Kong, Oct 12
  • 2. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 3. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 4. Recommender Systems (RS) • Information overload problem • Adapt user preference • Provide a list of recommendations • Assist decision makings E-Commerce (Amazon.com)
  • 5. Social RS (Twitter) Tagging RS (Flickr) Examples in other domains: Netflix, Pandora, Yelp, etc
  • 6. Affective Recommender Systems • Affective computing is the study and development of systems and devices that can recognize, interpret, process, and simulate human affects. It is an interdisciplinary field spanning computer sciences, psychology, and cognitive science. • Decision Making = Rational + Emotional (Daniel Kahnemann, Economic Nobel Prize 2002) • Affective Recommender System takes emotional effects into consideration to further assist decision makings and recommendations. E.g. music preferences by emotional effect M. Tkalcic, A. Kosir, J. Tasic. "Affective recommender systems: the role of emotions in recommender systems." Decisions@RecSys 2011
  • 7. Take Emotions into RS • Content-based Frameworks Semantic emotions can be extracted from tags (affective tags), reviews (review or opinion mining), user tweets, metadata, etc • Collaborative Filtering In a way of “collaborative”: see how other users prefer with similar emotions; or see how other items user may prefer within similar emotions • Context-aware Recommender Systems It is a multidimensional approach, where emotions are deemed as additional dimensions
  • 8. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 9. Context-aware RS (CARS) • Traditional RS: Users × Items  Ratings • CARS: Users × Items × Contexts  Ratings
  • 10. Context-aware RS (CARS) • Example: Tour Plan Recommender Yu, Chien-Chih and Chang, Hsiao-ping, "Towards Context-Aware Recommendation for Personalized Mobile Travel Planning". International Conference on Context-Aware Systems and Applications, ICCASA 2012
  • 11. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 12. The Role of Emotions in CARS • Several research has considered emotions as contexts and demonstrate the improved efficiency in contextual recommendation • But what is the role of emotions? Improved effectiveness in terms of MAE, RMSE, Precision, etc cannot fully reveal the roles.
  • 13. The Role of Emotions in CARS Definition: the role of emotions • Effectiveness: how effective the emotions are? Comparison among emotion only, no emotions, and a mixture of emotions & other features. • Usage: how emotions are used in CARS? E.g. which emotions are selected? which emotions are most influential ones? which components emotions are applied to? any latent patterns in terms of usage?
  • 14. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 15. Explore The Role of Emotions in CARS • Effectiveness: how effective the emotions are? Comparison among emotion only, no emotions, and a mixture of emotions & other features. • Usage: how emotions are used in CARS? E.g. which emotions are selected? which emotions are most influential ones? which components emotions are applied to? any latent patterns in terms of usage? • How to explore? Two classes of CARS algorithms: Context-aware Splitting Approaches; Differential Context Modeling; Baseline: Context-aware Matrix Factorization
  • 16. Context-aware Splitting Approaches • Three Algorithms: Item Splitting, User Splitting, UI Splitting • Item Splitting: The underlying idea is that the nature of an item, from the user's point of view, may change in different contextual conditions, hence it may be useful to consider it as two different items. (L. Baltrunas, F. Ricci, RecSys'09) At Cinema At Home At Swimming Pool
  • 17. Context-aware Splitting Approaches Splitting Process: Choose a contextual condition to split each item; The selection process is done by measuring significance of rating differences (such as the two-sample t test); Then multidimensional matrix is converted to a 2D one, then traditional algorithms like CF, MF can be applied to; User Item Loc Rating U1 M1 Pool 5 U2 M1 Pool 5 U3 M1 Pool 5 U1 M1 Home 2 U4 M1 Home 3 U2 M1 Cinema 2 High Ratings Low Ratings Significant difference? Let’s split it !!!
  • 18. Context-aware Splitting Approaches Matrix Transformation User Item Loc Rating U1 M1 Pool 5 U2 M1 Pool 5 U3 M1 Pool 5 U1 M1 Home 2 U4 M1 Home 3 U2 M1 Cinema 2 User Item Rating U1 M11 5 U2 M11 5 U3 M11 5 U1 M12 2 U4 M12 3 U2 M12 2 Transformation An alternative contextual condition for splitting: “Pool” vs. “Non-Pool” If there is qualified split, one item will be split to two new ones.  Contexts are fused into items as new items in 2D rating matrix;  Traditional RS algorithms can be applied to;
  • 19. Context-aware Splitting Approaches • User Splitting: is a similar one. Instead of splitting items, it may be useful to consider one user as two different users, if user demonstrates significantly different preferences across contexts. (L. Baltrunas, et al., CARS@RecSys 2009 and A. Said et al., CARS@RecSys 2011) • UI Splitting: simply a combination of item splitting and user splitting – both approaches are applied to create a new rating matrix – new users and new items are created in the rating matrix. (Y. Zheng, R. Burke, B. Mobasher, Decisions@RecSys 2013)
  • 21. Context-aware Splitting Approaches How about the characteristics of this approach in terms of discovering the usage of contexts in the recommendation process? Context-aware splitting approaches have to select influential contextual conditions to split users or items; so the usage can be inferred by the distribution of which emotional contexts are used to split items and users, and which ones are the most frequent used ones?
  • 22. Differential Context Modeling  Separate one algorithm to different components;  Apply optimal contextual constraints to each component. The underlying is that, one context may be influential for this component, but it is not usually the case for other components.
  • 23. Differential Context Modeling Contextual constraints can be applied by a condition-matching (differential context relaxation, DCR) or a similarity-matching (differential context weighting, DCW); In DCR, the contexts should exactly matched by selected variables. In DCW, a similar enough context is allowed – it is not necessary to be exactly matched, but should be similar enough. Location: Home, Cinema, Swimming Pool Time: Weekend, Weekday Companion: Family, Friend, Girlfriend If location and time are selected: DCR  exactly match {Cinema, Weekend} DCW{Home, Weekend} is 50% similar! (Simply, assume dimensions have equal weights in DCW)
  • 24. Differential Context Modeling How about the characteristics of this approach in terms of discovering the usage of contexts in the recommendation process? DCR  Which emotional dimensions are selected for each component? (feature selection) DCW  Which dimensions are weighted higher in each component? (feature weighting)
  • 25. Baseline in Performance: CAMF CAMF = Context-aware Matrix Factorization We just applied CAMF to compare predictive performance, and did not explore the usage of contexts in this approach. Basic MF: Bias MF: CAMF_CI: CAMF_CU: Reference: Baltrunas, Linas, Bernd Ludwig, and Francesco Ricci. "Matrix factorization techniques for context aware recommendation." Proceedings of the fifth ACM conference on Recommender systems. ACM, 2011. Three Shapes: CAMF_C, CAMF_CI, CAMF_CU
  • 26. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 27. Experimental Design Contextual variables in the Dataset: LDOS-CoMoDa Emotional contexts # of users: 113; # of items: 1186; # of ratings: 2094; rating scale: 1-5 Each user has at least 5 ratings profiles in the data;
  • 28. Experimental Design • Effectiveness: how effective the emotions are? Comparison among emotion only, no emotions, and a mixture of emotions & other features (all contexts). • Usage: how emotions are used in CARS? E.g. which emotions are selected? which emotions are most influential ones? which components emotions are applied to? any latent patterns in terms of usage?
  • 29. • Effectiveness: Comparison among emotion only, no emotions, and a mixture of emotions & other features (all contexts) by RMSE. Findings:  “No emotions” plays the worst; (Note: MF is applied to splitting approaches)  “Emotion Only” works the best for CAMF; “All contexts” is the best for other two approaches;  UI splitting achieves lowest RMSE in all situations: All contexts, no emotions & emotion only;  User splitting performs better than item splitting for this data;  Emotions are beneficial !!!
  • 30. Experimental Design • Effectiveness: how effective the emotions are? Comparison among emotion only, no emotions, and a mixture of emotions & other features (all contexts). • Usage: how emotions are used in CARS? E.g. which emotions are selected? which emotions are most influential ones? which components emotions are applied to? any latent patterns in terms of usage?  Context-aware splitting approaches the usage distribution of contexts used to split items or users  Differential context modeling DCR  Which dimensions are selected for each component? DCW  Which dimensions are weighted higher in each component?
  • 31. • Usage by Context-aware Splitting Approaches Top-3 selections in item splitting: EndEmotion, Time, Season Top-3 selections in user splitting: EndEmotion, DominantEmo, Time Note: User splitting outperforms item splitting, any patterns? Conjecture: dependences between contexts and users|items.
  • 32. Experimental Results • Usage by Differential Context Modeling Notice: we applied DCM to user-based Collaborative Filtering.  DCR  Which contexts are selected for each component?  DCW  Which contexts are weighted higher in each component? For DCW, we did not show weights, but give selected dimensions weighted higher than 0.7. In DCR, emotional variables are selected in components of user baseline and similarity; In DCW, emotions are weighted significantly higher expect for neighbor contribution; DCM is optimal algorithm in context selection and weighting. Once emotions are selected or weighted highly, they are optimal solutions!
  • 33. • Summary of the Usage Patterns In Context-aware Splitting Approaches:  EndEmo and DominantEmo are two most influential ones!  And emotion variables defeat other variables! (topper selection!)  User splitting works better than Item splitting for this data In Differential Context Modeling:  EndEmo and DominantEmo are selected in DCR, and all three emotions are selected in DCW.  EndEmo and DominantEmo are influential for two components in both DCR and DCW: user baseline and user similarity calculation Overlapped findings: EndEmo and DominantEmo are influential ones!
  • 34. • Recommender Systems (RS) & Affective RS • Context-aware Recommender Systems (CARS) • The Role of Emotions in CARS • Explore The Roles by CARS Algorithms • Experimental Results • Conclusions and Future Work
  • 35. • What we did? Emotions are demonstrated to improve recommendation performance, but few research focus on the other side of the roles of emotions – the usage instead of performance only. In this work, we explore the roles (effectiveness + usage) by two classes of CARS algorithm: context- aware splitting approach, and differential context modeling • What we found? Effectiveness: UI splitting performs the best among all CARS algorithm for this data set; A mixture of all contexts is better for splitting approaches and DCM; CAMF may require context selection in advance; Usage: Splitting methods can explore usage by distribution of how frequent one context is used to split items or users; DCM infers the selection or the degree of importance by weights for each algorithm component; It can tell the importance of contexts by algorithm components. Patterns: Among three emotional variables, DominantEmo & EndEmo are two influential ones; Conjecture: User splitting may outperform item splitting if contexts are more dependent with users than items; experiments on more data are required to confirm this pattern.
  • 36. Future Work  The emotions in this data are not real emotions, e.g. DominantEmo and EndEmo are post-emotions, how about pre-emotions (emotion before the activity)?  What are the insight of those emotions? Any associations among emotions, user profiles and item features? For example, EndEmo is sad after seeing a tragedy, but rating is high (because tragedy makes people cry…), Or, a happy EndEmo will always results in higher rating? (because users likes this experience).  In terms of the three splitting approaches, especially the new proposed one – UI Splitting, it is useful to perform an empirical study among those three splitting approaches to further explore in which situation, which one will perform better than other ones.
  • 37. References Item Splitting 1) L. Baltrunas, and F. Ricci. "Context-based splitting of item ratings in collaborative filtering." RecSys, 2009. 2) L. Baltrunas, and F. Ricci. "Experimental evaluation of context-dependent collaborative filtering using item splitting." User Modeling and User-Adapted Interaction (2013): 1-28. User Splitting 1) L. Batrunas and X. Amatriain."Towards Time-Dependent Recommendation Based on Implicit Feedback." CARS@RecSys, 2009 2) A. Said, E. Luca, S. Albayrak. "Inferring contextual user profiles—improving recommender performance.“ CARS@RecSys, 2011 UI Splitting 1) Y. Zheng, R. Burke, B. Mobasher. "The Role of Emotions in Context-aware Recommendation". Decisons@RecSys, 2013 Context-aware Matrix Factorization 1) L. Baltrunas, B. Ludwig, F. Ricci. "Matrix factorization techniques for context aware recommendation." RecSys 2011. Differential Context Modeling 1) Y. Zheng, R. Burke, B. Mobasher. "Differential Context Relaxation for Context-aware Travel Recommendation". EC-WEB, 2012 2) Y. Zheng, R. Burke, B. Mobasher. "Optimal Feature Selection for Context-Aware Recommendation using Differential Relaxation". CARS@RecSys, 2012 3) Y. Zheng, R. Burke, B. Mobasher. "Recommendation with Differential Context Weighting". In UMAP, 2013
  • 38. Thank You! Decisions@RecSys 2013, Hong Kong Center for Web Intelligence, DePaul University, USA 10/12/2013