SlideShare une entreprise Scribd logo
1  sur  39
“I like to explore sometimes”:
Adapting to Dynamic User
Novelty Preferences
Komal
Kapoor
Vikas
Kumar
Joe
Konstan
Paul
Schrater
Loren
Terveen
1Twitter: #adaNovR
In this paper:
predict novelty preferences based on past
consumption behavior.
users’ novelty preferences vary over time.
adaptive novelty recommender
2
What are Preferences?
Preferences determine how we make our
everyday choices…
3
What are Preferences?
Preferences determine how we make our
everyday choices…
Wait!!
Everyone knows this!
4
What are Preferences?
Novelty Preferences determine how well we
appreciate new items..
Novelty
Source: tumblr
5
Why we want to understand these
novelty preferences?
Explore Recommend
same old
Bored Churn
Exploit Recommend
New Items
Frustration Churn
6
Why we want to understand these
novelty preferences?
Explore Recommend
same old
Bored Churn
Exploit Recommend
New Items
Frustration Churn
7
Static Preference Models
Similar Users Have
Similar Preferences
Users Prefer
Similar Items
User-based Filtering Item-based Filtering
8
Static Preference Models
Similar Users Have
Similar Preferences
Users Prefer
Similar Items
User-based Filtering Item-based Filtering
No understanding of user
consumption behavior.
Fails when preferences change!!
9
Dynamic Novelty Preference
» not every user seek new
items
» some users seek (or
explore) more
» even they do it sometimes
10
Dynamic Novelty Preference
» not every user seek new
items
» some users seek (or
explore) more
» even they do it sometimes
11
Add value to user experience by
understanding their (changing) need
better
Data
» Music data:
• Closely related to human emotions and behavior
responses
• Low risk/cost of consumption
» Two Datasets:
NDA
12
Details: last.fm
last.fm
Duration 3 months
Number of Users 882
Avg. #session per user 56
Avg. session Length
(#items)
39
13
User Timeline: Sessions
14
User Timeline: Definitions
• Familiar Set:
• items recently consumed by user (within time
window T)
• Novel or New Set:
• New items consumed **compared to
previous familiar set (T-1)**
15
User Timeline: Definitions
• Familiar Set:
• items recently consumed by user (within time
window T)
• Novel or New Set:
• New items consumed **compared to
previous familiar set** set.
Novelty Seeking Score (nvSeek) =
#new-items / #unique-items
16
In this paper:
predict novelty preferences based on past
consumption behavior.
users’ novelty preferences vary over time.
adaptive novelty recommender
17
Results (1):
Users have different novelty preferences
Novelty Seeking Score
NumberofUsers
• We have some high
as well as some low
novelty seeking
users.
• Scores vary across
the users ( s.d >
0, p-val ~ 0)
18
Results (2):
Users have dynamic novelty preferences
Seeking Deviation
NumberofUsers
• users’ seeking score
deviation across multiple
time window.
• users show dynamic
seeking score over a
period of three months
(Mean > 0, pval ~ 0)
19
In this paper:
predict novelty preferences based on past
consumption behavior.
users’ novelty preferences vary over time.
adaptive novelty recommender
20
Intuition
Diverse users are likely to be
more novelty seeking
Diversity of the familiar set
User bored with their current
selection are likely to be more
novelty seeking
Boredom with the familiar set
21
Model features:
» Diversity = more items, more diverse
» Boredom:
• Dynamic Item Preference [Kapoor et al, WSDM 2015]
– More you play, OR
– less gap between your plays
Fast to reach boredom
22
Model
» Logistic Regression Model:
• prediction novelty preference score based on
past consumptions
23
Results:
» Accurate seeking score predictions than
constant novelty.
» Both features are significant and
positively correlated
• higher diversity seeking new items
• higher boredom seeking new items
24
In this paper:
predict novelty preferences based on past
consumption behavior.
users’ novelty preferences vary over time.
adaptive novelty recommender
25
Adaptive Recommendation
» Existing Systems:
» Adaptive System:
Novelty
Seeking
Score
26
Adaptive Novelty Recommendation
Novelty
Preference
High!
F1
F2
F3
F4
….
N1
N2
N3
N4
….
F1
N1
N2
F2
N3
….
Novelty
Preference
Low!
F1
F2
F3
F4
….
N1
N2
N3
N4
….
F1
F2
F2
N1
F4
….
27
Design
Adaptive Recommendation
Module
Novelty Seeking Prediction Module
Item Ranking Module
User Timelines
(Past Sessions)
Ranked
Familiar
Items
Ranked
Novel
Items
Consumed/
Rated Items
Behavioral/
Session
Attributes
Novelty
Seeking
Input Output
OutputInput
Top-N
Recommendations
- Item1
- Item2
- ItemN
Current Session
Recommendations
System Design
Adaptive
Ranking
Explicit
Implicit
28
Evaluation
» Baselines
• Item Based CF
• Constant Novelty
• PureN : only novel or new items
• PureF : only familiar items
29
Metrics
» Metrics
• Recommendation Accuracy
• cost sensitive weighted F-measure
• Novelty Accuracy:
• new items recommended Vs. new items consumed
30
Results
Recommendation Accurracy
Last.fm
Weighted F-measure for different novelty seeking score
31
Results
Recommendation Accurracy
Last.fm
Weighted F-measure for different novelty seeking score
Performance of PureF and Item Based
declines as novelty seeking score
increases
32
Results
Recommendation Accurracy
Last.fm
Weighted F-measure for different novelty seeking score
adaNov-R performs comparable to the
best baseline for all novelty seeking
scores 33
Novelty Accuracy
34
Novelty Accuracy
adaNov-R capable to adapt to number of
new items.
35
Novelty Accuracy
PureN provides all new items.
Item based CF rarely provides new
items
36
Key Takeaways:
» Novelty Preferences are dynamic across
and within users
» Past consumption provides significant
signal to predict future novelty
preferences.
» A recommender capable to adapt to
novelty preference
37
Conclusion
» Modeling novelty preference dynamics
significantly impacts recommendation
design
» Future Work:
• Study the effect on retention due to adaptive
recommendations.
38
Danke!!
(thanks!!)
Supported by National Science Foundation under grants IIS 08-08692, IIS 09-
64695, UMN SOBACO grant and Doctoral Dissertation Fellowship.
Questions?
39

Contenu connexe

Similaire à "I like to explore sometimes": Adapting to Dynamic User Novelty Preferences

Best data science courses in pune
Best data science courses in puneBest data science courses in pune
Best data science courses in puneprathyusha1234
 
Top data science institutes in hyderabad
Top data science institutes in hyderabadTop data science institutes in hyderabad
Top data science institutes in hyderabadprathyusha1234
 
best online data science courses
best online data science coursesbest online data science courses
best online data science coursesprathyusha1234
 
Recommender systems
Recommender systemsRecommender systems
Recommender systemsTamer Rezk
 
case based recommendation approach for market basket data
case based recommendation approach for market basket datacase based recommendation approach for market basket data
case based recommendation approach for market basket datamniranjanmurthy
 
Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender systemStanley Wang
 
Usability evaluations (part 2)
Usability evaluations (part 2) Usability evaluations (part 2)
Usability evaluations (part 2) Andres Baravalle
 
Introduction to recommender systems
Introduction to recommender systemsIntroduction to recommender systems
Introduction to recommender systemsAndrea Gigli
 
The Universal Recommender
The Universal RecommenderThe Universal Recommender
The Universal RecommenderPat Ferrel
 
Tutorial on sequence aware recommender systems - UMAP 2018
Tutorial on sequence aware recommender systems - UMAP 2018Tutorial on sequence aware recommender systems - UMAP 2018
Tutorial on sequence aware recommender systems - UMAP 2018Paolo Cremonesi
 
Bmgt 311 chapter_4
Bmgt 311 chapter_4Bmgt 311 chapter_4
Bmgt 311 chapter_4Chris Lovett
 
Recommender systems for E-commerce
Recommender systems for E-commerceRecommender systems for E-commerce
Recommender systems for E-commerceAlexander Konduforov
 
Recommendation engines
Recommendation enginesRecommendation engines
Recommendation enginesGeorgian Micsa
 
Mod5_Recommendation Systems.pptx
Mod5_Recommendation Systems.pptxMod5_Recommendation Systems.pptx
Mod5_Recommendation Systems.pptxdivyammo
 

Similaire à "I like to explore sometimes": Adapting to Dynamic User Novelty Preferences (20)

Best data science courses in pune
Best data science courses in puneBest data science courses in pune
Best data science courses in pune
 
Top data science institutes in hyderabad
Top data science institutes in hyderabadTop data science institutes in hyderabad
Top data science institutes in hyderabad
 
best online data science courses
best online data science coursesbest online data science courses
best online data science courses
 
Recommender systems
Recommender systemsRecommender systems
Recommender systems
 
case based recommendation approach for market basket data
case based recommendation approach for market basket datacase based recommendation approach for market basket data
case based recommendation approach for market basket data
 
Overview of recommender system
Overview of recommender systemOverview of recommender system
Overview of recommender system
 
Recommenders.ppt
Recommenders.pptRecommenders.ppt
Recommenders.ppt
 
Recommenders.ppt
Recommenders.pptRecommenders.ppt
Recommenders.ppt
 
Usability evaluations (part 2)
Usability evaluations (part 2) Usability evaluations (part 2)
Usability evaluations (part 2)
 
Bmgt 311 week_4
Bmgt 311 week_4Bmgt 311 week_4
Bmgt 311 week_4
 
Introduction to recommender systems
Introduction to recommender systemsIntroduction to recommender systems
Introduction to recommender systems
 
The Universal Recommender
The Universal RecommenderThe Universal Recommender
The Universal Recommender
 
Tutorial on sequence aware recommender systems - UMAP 2018
Tutorial on sequence aware recommender systems - UMAP 2018Tutorial on sequence aware recommender systems - UMAP 2018
Tutorial on sequence aware recommender systems - UMAP 2018
 
Bmgt 311 chapter_4
Bmgt 311 chapter_4Bmgt 311 chapter_4
Bmgt 311 chapter_4
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Recommender systems for E-commerce
Recommender systems for E-commerceRecommender systems for E-commerce
Recommender systems for E-commerce
 
Recommendation engines
Recommendation enginesRecommendation engines
Recommendation engines
 
MORS22.pdf
MORS22.pdfMORS22.pdf
MORS22.pdf
 
Mod5_Recommendation Systems.pptx
Mod5_Recommendation Systems.pptxMod5_Recommendation Systems.pptx
Mod5_Recommendation Systems.pptx
 
Recommender systems
Recommender systemsRecommender systems
Recommender systems
 

Dernier

9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000Sapana Sha
 
Zoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdfZoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdfSumit Kumar yadav
 
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verifiedConnaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verifiedDelhi Call girls
 
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bAsymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bSérgio Sacani
 
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...Silpa
 
module for grade 9 for distance learning
module for grade 9 for distance learningmodule for grade 9 for distance learning
module for grade 9 for distance learninglevieagacer
 
GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)Areesha Ahmad
 
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑Damini Dixit
 
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Servicenishacall1
 
Forensic Biology & Its biological significance.pdf
Forensic Biology & Its biological significance.pdfForensic Biology & Its biological significance.pdf
Forensic Biology & Its biological significance.pdfrohankumarsinghrore1
 
Conjugation, transduction and transformation
Conjugation, transduction and transformationConjugation, transduction and transformation
Conjugation, transduction and transformationAreesha Ahmad
 
PSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptxPSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptxSuji236384
 
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICESAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICEayushi9330
 
Digital Dentistry.Digital Dentistryvv.pptx
Digital Dentistry.Digital Dentistryvv.pptxDigital Dentistry.Digital Dentistryvv.pptx
Digital Dentistry.Digital Dentistryvv.pptxMohamedFarag457087
 
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticsPulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticssakshisoni2385
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Monika Rani
 
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Service
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts ServiceJustdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Service
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Servicemonikaservice1
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and ClassificationsAreesha Ahmad
 

Dernier (20)

9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 60009654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
9654467111 Call Girls In Raj Nagar Delhi Short 1500 Night 6000
 
Zoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdfZoology 5th semester notes( Sumit_yadav).pdf
Zoology 5th semester notes( Sumit_yadav).pdf
 
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verifiedConnaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
Connaught Place, Delhi Call girls :8448380779 Model Escorts | 100% verified
 
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 bAsymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
Asymmetry in the atmosphere of the ultra-hot Jupiter WASP-76 b
 
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
+971581248768>> SAFE AND ORIGINAL ABORTION PILLS FOR SALE IN DUBAI AND ABUDHA...
 
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...
Locating and isolating a gene, FISH, GISH, Chromosome walking and jumping, te...
 
module for grade 9 for distance learning
module for grade 9 for distance learningmodule for grade 9 for distance learning
module for grade 9 for distance learning
 
GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)GBSN - Biochemistry (Unit 1)
GBSN - Biochemistry (Unit 1)
 
Site Acceptance Test .
Site Acceptance Test                    .Site Acceptance Test                    .
Site Acceptance Test .
 
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
High Profile 🔝 8250077686 📞 Call Girls Service in GTB Nagar🍑
 
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
9999266834 Call Girls In Noida Sector 22 (Delhi) Call Girl Service
 
Forensic Biology & Its biological significance.pdf
Forensic Biology & Its biological significance.pdfForensic Biology & Its biological significance.pdf
Forensic Biology & Its biological significance.pdf
 
Conjugation, transduction and transformation
Conjugation, transduction and transformationConjugation, transduction and transformation
Conjugation, transduction and transformation
 
PSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptxPSYCHOSOCIAL NEEDS. in nursing II sem pptx
PSYCHOSOCIAL NEEDS. in nursing II sem pptx
 
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICESAMASTIPUR CALL GIRL 7857803690  LOW PRICE  ESCORT SERVICE
SAMASTIPUR CALL GIRL 7857803690 LOW PRICE ESCORT SERVICE
 
Digital Dentistry.Digital Dentistryvv.pptx
Digital Dentistry.Digital Dentistryvv.pptxDigital Dentistry.Digital Dentistryvv.pptx
Digital Dentistry.Digital Dentistryvv.pptx
 
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceuticsPulmonary drug delivery system M.pharm -2nd sem P'ceutics
Pulmonary drug delivery system M.pharm -2nd sem P'ceutics
 
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
Vip profile Call Girls In Lonavala 9748763073 For Genuine Sex Service At Just...
 
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Service
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts ServiceJustdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Service
Justdial Call Girls In Indirapuram, Ghaziabad, 8800357707 Escorts Service
 
Bacterial Identification and Classifications
Bacterial Identification and ClassificationsBacterial Identification and Classifications
Bacterial Identification and Classifications
 

"I like to explore sometimes": Adapting to Dynamic User Novelty Preferences

Notes de l'éditeur

  1. ----- Meeting Notes (9/16/15 14:43) ----- Explore Vs Exploit!! Lean back users Vs no interruption required users - One step closer to understand user behavior and able to make require changes in recommendations - Team!! (PhD Candidate @ GroupLens at University of Minnesota) - Paul - pysychological perspective
  2. Most of you listen to music. Raise your hand if you have been listening to same music playlist for over a month now? Now, raise your hand if you try to change your playlist often searching for new or music from the past?
  3. Most of you listen to music. Raise your hand if you have been listening to same music playlist for over a month now? Now, raise your hand if you try to change your playlist often searching for new or music from the past?
  4. Most of you listen to music. Raise your hand if you have been listening to same music playlist for over a month now? Now, raise your hand if you try to change your playlist often searching for new or music from the past?
  5. Now, if you
  6. We confirm the results with the other data too. Details are in the paper