integra(ng)social)with)search) - stanford...
TRANSCRIPT
![Page 1: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/1.jpg)
Integra(ng Social with Search
Edward Chang
Director, Google Research, Beijing�
WISE Keynote 1 2010-12-13
![Page 2: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/2.jpg)
Related Papers �• AdHeat (Social Ads):
– AdHeat: An Influence-‐based Diffusion Model for PropagaAng Hints to Match Ads, H.J. Bao and E. Y. Chang, WWW 2010 (best paper candidate), April 2010.
– Parallel Spectral Clustering in Distributed Systems, Wen-‐Yen Chen, Yangqiu Song, Hongjie Bai, Chih-‐Jen Lin, and E. Y. Chang, IEEE TransacAons on PaTern Analysis and Machine Intelligence (PAMI), 2010.
• UserRank: – Confucius and its Intelligent Disciples, X. Si, E. Y. Chang, Z. Gyongyi, VLDB, September 2010 .
– Topic-‐dependent User Rank, Xiance Si, Z. Gyongyi, E. Y. Chang, and M.S. Sun, Google Technical Report.
• Large-‐scale Collabora;ve Filtering: – PLDA+: Parallel Latent Dirichlet AllocaAon for Large-‐Scale ApplicaAons, ACM TransacAons on
Internet Technology, 2011. – CollaboraAve Filtering for Orkut CommuniAes: Discovery of User Latent Behavior, W.-‐Y. Chen,
J. Chu, E. Y. Chang, WWW 2009: 681-‐690. – CombinaAonal CollaboraAve Filtering for Personalized Community RecommendaAon, W.-‐Y.
Chen, E. Y. Chang, KDD 2008: 115-‐123. – PSVM: Parallelizing SVMs on distributed machines, E. Y. Chang, et al., NIPS 2007.
WISE Keynote 2 2010-12-13
![Page 3: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/3.jpg)
Web 1.0
WISE Keynote 3
.htm
.htm
.htm
.jpg
.jpg
.doc
.htm
.msg
.htm
.htm
2010-12-13
![Page 4: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/4.jpg)
Web 2.0 -‐-‐-‐ Web with People
WISE Keynote 4
.htm
.jpg
.doc
.xls
.msg .htm
.htm
.jpg
.msg
.htm
2010-12-13
![Page 5: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/5.jpg)
WISE Keynote 5 2010-12-13
![Page 6: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/6.jpg)
WISE Keynote 6 2010-12-13
![Page 7: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/7.jpg)
Outline�
• Search + Social Synergy • Search Social
• Social Search
• Scalability�
WISE Keynote 7 2010-12-13
![Page 8: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/8.jpg)
Google Q&A (Confucius)
• Developed from 2007 All now @ Beijing • Launched in more than 60 courAers
– Russia – HK – Southeast Asia – Arab World
– Sub-‐Saharan Africa (Baraza)
WISE Keynote 8 2010-12-13
![Page 9: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/9.jpg)
WISE Keynote 9
Query: What are must-‐see a<rac(ons at Yellowstone
2010-12-13
![Page 10: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/10.jpg)
WISE Keynote 10
Query: What are must-‐see a<rac(ons at Yellowstone
2010-12-13
![Page 11: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/11.jpg)
WISE Keynote 11
Query: What are must-‐see a<rac(ons at Yosemite
2010-12-13
![Page 12: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/12.jpg)
WISE Keynote 12
Query: What are must-‐see a<rac(ons at Beijing
Hotel ads
2010-12-13
![Page 13: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/13.jpg)
WISE Keynote 13
Search Quality at Stake.
Strategic Issue
13 17 June 2008 -‐ Knowledge Search Strategy
% of Referral Traffic From 1st Page Sent to Yahoo / Baidu % of First Result Page with >=1 Q&A Result from Yahoo or Baidu
61 countries have Q&A or advanced forums as top 10 most clicked des;na;on (out of 115 countries with more than 1M session)
2010-12-13
![Page 14: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/14.jpg)
SNS & Mobile Also Need Q&A
• Social Networks – Difficult to find user intent to match ads
– Q&A is a perfect app to learn users’ problems
• Mobile Search – Voice is the most convenient user interface – Succinct search result (or rich snippets) is desirable
WISE Keynote 14 2010-12-13
![Page 15: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/15.jpg)
Confucius: Google Q&A Providing High-‐Quality Answers in a Timely Fashion
Trigger a discussion/quesAon session during search Provide labels to a post (semi-‐automaAcally) Given a post, find similar posts (automaAcally) Evaluate quality of a post, relevance and originality Evaluate user credenAals in a topic sensiAve way Route quesAons to experts Provide most relevant, high-‐quality content for Search to index Generate answers using NLP
WISE Keynote 15
U
Search
S U
Community
C
CCS/Q&A
Discussion/Ques;on
DC
Differen;ated Content
2010-12-13
![Page 16: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/16.jpg)
Confucius: Google Q&A Providing High-‐Quality Answers in a Timely Fashion
Trigger a discussion/quesAon session during search Provide labels to a quesAon (semi-‐automaAcally) Given a quesAon, find similar quesAons (automaAcally) Evaluate quality of an answer, relevance and originality Evaluate user credenAals in a topic sensiAve way Route quesAons to experts Provide most relevant, high-‐quality content for Search to index Generate answers using NLP
WISE Keynote 16
U
Search
S U
Community
C
CCS/Q&A
Discussion/Ques;on
DC
Differen;ated Content
2010-12-13
![Page 17: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/17.jpg)
WISE Keynote 17
Q&A Uses Machine Learning
Label suggestion using LDA algorithm.
• Real Time topic-to-topic (T2T) recommendation using LDA algorithm.
• Gives out related high quality links to previous questions before human answer appear.
2010-12-13
![Page 18: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/18.jpg)
WISE Keynote 18
1 1 1 1 1 1 1
1 1 1 1
1 1 1 1
1 1 1 1 1
1 1 1 1 1 1 1 1 1
1 1 1
Que
stio
ns
Labels/Qs
Based on membership so far, and memberships of others
Predict further membership
CollaboraAve Filtering
2010-12-13
![Page 19: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/19.jpg)
FIM-‐based RecommendaAon
WISE Keynote 19 2010-12-13
![Page 20: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/20.jpg)
Distributed Latent Dirichlet AllocaAon (LDA)
WISE Keynote 20
• Other CollaboraAve Filtering Apps – Recommend Users Users – Recommend Music Users – Recommend Ads Users – Recommend Answers Q
• Predict the ? In the light-‐blue cells
Users/Music/Ads/Question
Use
rs/M
usic
/Ads
/Ans
wer
s
1 recipe pastry for a 9 inch double crust 9 apples, 2/1 cup, brown sugar
How to install apps on Apple mobile phones?
Documents
Topic Distribution
Topic Distribution
User quries iPhone crack Apple pie
• Search – Construct a latent layer for beTer
for semanAc matching
• Example: – iPhone crack – Apple pie
2010-12-13
![Page 21: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/21.jpg)
Latent Dirichlet AllocaAon [D. Blei, M. Jordan 04]
• α: uniform Dirichlet φ prior for per document d topic distribuAon (corpus level parameter)
• β: uniform Dirichlet φ prior for per topic z word distribuAon (corpus level parameter)
• θd is the topic distribuAon of document d (document level)
• zdj the topic if the jth word in d, wdj the specific word (word level)
WISE Keynote 21
θ
z
w
Nm
M
α
βφ
K
2010-12-13
![Page 22: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/22.jpg)
CombinaAonal CollaboraAve Filtering Model (CCF) [W.-‐Y. Chen, et al, KDD2008]
WISE Keynote 22
Communities
users descriptions users descriptions
Communities Communities
Word 1
Word 2
Word 3
2010-12-13
![Page 23: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/23.jpg)
WISE Keynote 23 2010-12-13
![Page 24: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/24.jpg)
Confucius: Google Q&A
Trigger a discussion/quesAon session during search Provide labels to a post (semi-‐automaAcally) Given a post, find similar posts (automaAcally) Evaluate quality of a post, relevance and originality Evaluate user credenAals in a topic sensiAve way Route quesAons to experts Provide most relevant, high-‐quality content for Search to index NLQA
WISE Keynote 24
U
Search
S U
Community
C
CCS/Q&A
Discussion/Ques;on
DC
Differen;ated Content
2010-12-13
![Page 25: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/25.jpg)
UserRank
WISE Keynote 25
• Rank users by quanAty (number of links) and quality (weights on links) of contribuAons
• Quality include: – Relevance. Is an answer relevant to the
Q? Measured by KL divergence between latent-‐topic vectors of A and Q
– Coverage. Compared among different answers
– Originality. Detect potenAal plagiarism and spam
– Promptness. Time between Q and A posAng Ame
2010-12-13
![Page 26: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/26.jpg)
Outline�
• Search + Social Synergy • Social Search
• Search Social
• Scalability �
WISE Keynote 26 2010-12-13
![Page 27: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/27.jpg)
Outline�
• Search + Social Synergy • Social Search
• Search Social
• Scalability�
WISE Keynote 27 2010-12-13
![Page 28: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/28.jpg)
Social?
• ConnecAng to friends • Knowing what friends are up to • ConnecAng to strangers
– DaAng, Gaming – Shopping
• Making recommendaAons based on acAviAes
WISE Keynote 28 2010-12-13
![Page 29: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/29.jpg)
User Latent Model
• α: uniform Dirichlet φ prior for per user u interest distribuAon (populaAon level parameter)
• β: uniform Dirichlet φ prior for per interest z acAvity distribuAon (populaAon level parameter)
• θd is the interest distribuAon of user u (user level)
• zuj the interest of the jth acAvity in u, wuj the specific acAvity (acAvity level)
WISE Keynote 29
θ
z
w
Nm
M
α
βφ
K
2010-12-13
![Page 30: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/30.jpg)
WISE Keynote 30 2010-12-13
![Page 31: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/31.jpg)
CombinaAonal CollaboraAve Filtering Model (CCF)
WISE Keynote 31
Users
Keywords Videos/Photos Keywords Videos/Photos
Users Users
2010-12-13
![Page 32: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/32.jpg)
Interest Networks
WISE Keynote 32 2010-12-13
![Page 33: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/33.jpg)
Outline�
• Search + Social Synergy • Social Search
– Mobilize users to improve search quality – Google Q&A, Facebook Like
• Search Social – Use query log to help social
• AcAviAes Interests Social • Groupcom
• Scalability�
WISE Keynote 33 2010-12-13
![Page 34: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/34.jpg)
34 WISE Keynote 2010-12-13
![Page 35: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/35.jpg)
35 WISE Keynote 2010-12-13
![Page 36: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/36.jpg)
36 WISE Keynote 2010-12-13
![Page 37: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/37.jpg)
37 WISE Keynote 2010-12-13
![Page 38: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/38.jpg)
38 WISE Keynote 2010-12-13
![Page 39: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/39.jpg)
39 WISE Keynote 2010-12-13
![Page 40: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/40.jpg)
40 WISE Keynote 2010-12-13
![Page 41: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/41.jpg)
User Latent Model
• α: uniform Dirichlet φ prior for per user u interest distribuAon (populaAon level parameter)
• β: uniform Dirichlet φ prior for per interest z acAvity distribuAon (populaAon level parameter)
• θd is the interest distribuAon of user u (user level)
• zuj the interest of the jth acAvity in u, wuj the specific acAvity (acAvity level)
WISE Keynote 41
θ
z
w
Nm
M
α
βφ
K
2010-12-13
![Page 42: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/42.jpg)
LDA Gibbs Sampling: Inputs & Outputs
WISE Keynote 42
users
words topics
words topics
users
Inputs: 1. training data: users as bags of
words 2. parameter: the number of
topics
Outputs: 1. model parameters: a co-‐
occurrence matrix of topics and words.
2. by-‐product: a co-‐occurrence matrix of topics and users.
2010-12-13
![Page 43: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/43.jpg)
Parallel Gibbs Sampling
WISE Keynote 43
Inputs: 1. training data: users as bags of
words 2. parameter: the number of
topics
Outputs: 1. model parameters: a co-‐
occurrence matrix of topics and words.
2. by-‐product: a co-‐occurrence matrix of topics and users.
users
words topics
words topics
users
2010-12-13
![Page 44: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/44.jpg)
ObservaAons
WISE Keynote 44
Master Node
• BoTleneck: CommunicaAon
• Amdahl’s law caps speedup
• Words in a bag have no order
• Words on a computer node can be reordered
2010-12-13
![Page 45: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/45.jpg)
Example Bags / Node A
• Bag #1 w1, w2, w3, w1, w2, w3, w1, w2, w3 • Bag #2 w1, w2, w1, w2, w1, w2, w1, w2 • Bag #3 w3, w1, w3, w1, w3, w1, w3, w1
• Bundle #1 w1, w1, w1, w1, w1, w1, … • Bundle #2 w2, w2, w2,… , • Bundle #3 w3, w3, w3,…
WISE Keynote 45 2010-12-13
![Page 46: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/46.jpg)
Two Nodes
Node A Node B
W1 W2
W2 W3
W3 W1
WISE Keynote 46 2010-12-13
![Page 47: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/47.jpg)
Parallel Gibbs Sampling
WISE Keynote 47
Inputs: 1. training data: documents as
bags of words 2. parameter: the number of
topics
Outputs: 1. model parameters: a co-‐
occurrence matrix of topics and words.
2. by-‐product: a co-‐occurrence matrix of topics and documents.
docs
words topics
words topics
docs
2010-12-13
![Page 48: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/48.jpg)
PLDA -‐-‐ enhanced parallel LDA�
• Take advantage of bag of words modeling: each Pw machine processes vocabulary in a word order
• Pipelining: fetching the updated topic distribuAon matrix while doing Gibbs sampling
WISE Keynote �48 2010-12-13
![Page 49: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/49.jpg)
Speedup 1,500x using 2,000 machines �
WISE Keynote 49 2010-12-13
![Page 50: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/50.jpg)
Lessons Learned
• BoTleneck MaTers • Inter-‐iteraAon MaTers
WISE Keynote 50 2010-12-13
![Page 51: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/51.jpg)
MapReduce�
WISE Keynote 51
map
map
map
map
map
Reduce
Data Block 1
Data Block 2
Data Block 3
Data Block 4
Data Block 5
Data datadatadatadatadatadatadatadatadatadata
Results datadatadatadatadatadata
2010-12-13
![Page 52: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/52.jpg)
Parallel Programming Models
WISE Keynote 52
MapReduce Project + MPI
GFS/IO and task rescheduling overhead between iterations
Yes No +1
No +1
Flexibility of computation model AllReduce only +0.5
? +1
Flexible +1
Efficient AllReduce Yes +1
Yes +1
Yes +1
Recover from faults between iterations
Yes +1
Yes +1
Apps
Recover from faults within each iteration
Yes +1
Yes +1
Apps
Final Score for scalable machine learning
3.5 5 4
2010-12-13
![Page 53: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/53.jpg)
53
SVM BoTlenecks Time consuming – 1M dataset, 8 days
Memory consuming – 1M dataset, 10G
... ... ..
.
WISE Keynote 2010-12-13
![Page 54: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/54.jpg)
54
Matrix FactorizaAon AlternaAves
exact
approximate
WISE Keynote 2010-12-13
![Page 55: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/55.jpg)
55
Raw Data Matrix Multiplication ICF
Incremental Data
Kernel Matrix
Incremental Kernel Matrix
Incremental ICF
Incremental Matrix Multiplication
Incremental Linear System Solving
Parallelizing SVM [E. Chang, et al, NIPS 07]
WISE Keynote 2010-12-13
![Page 56: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/56.jpg)
56
Incomplete Cholesky FactorizaAon (ICF)
n x n n x p p x n
WISE Keynote 2010-12-13
![Page 57: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/57.jpg)
57
Raw Data Matrix Multiplication ICF
Incremental Data
Kernel Matrix
Incremental Kernel Matrix
Incremental ICF
Incremental Matrix Multiplication
Incremental Linear System Solving
PSVM
WISE Keynote 2010-12-13
![Page 58: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/58.jpg)
58
Matrix Product
=
p x n n x p p x p
WISE Keynote 2010-12-13
![Page 59: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/59.jpg)
59
PSVM [E. Chang, et al, NIPS 07]
• Column-‐based ICF – Slower than row-‐based on single machine
– Parallelizable on mulAple machines
• Changing IPM computaAon order to achieve parallelizaAon
WISE Keynote 2010-12-13
![Page 60: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/60.jpg)
60
Overheads
WISE Keynote 2010-12-13
![Page 61: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/61.jpg)
61
Speedup
WISE Keynote 2010-12-13
![Page 62: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/62.jpg)
Scalability
• ComputaAon – ParallelizaAon – ApproximaAon
• File Systems – Latency – Recovery
• Power Management
WISE Keynote 62 2010-12-13
![Page 63: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/63.jpg)
Sample Plaxorms �
WISE Keynote 63 2010-12-13
![Page 64: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/64.jpg)
Sample Hierarchy�
• Server – 16GB DRAM; 160MB Flash; 5 x 1TB disk
• Rack – 40 servers – 48 port Gigabit Ethernet switch
• Warehouse – 10,000 servers (250 racks) – 2K port Gigabit Ethernet switch �
64 WISE Keynote 2010-12-13
![Page 65: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/65.jpg)
Storage -‐-‐-‐ One Server�
65 WISE Keynote 2010-12-13
![Page 66: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/66.jpg)
Storage -‐-‐-‐ One Rack�
66 WISE Keynote 2010-12-13
![Page 67: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/67.jpg)
Storage -‐-‐-‐ One Center�
67 WISE Keynote 2010-12-13
![Page 68: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/68.jpg)
Google File System (GFS)
• Master manages metadata • Data transfers happen directly between clients/chunkservers • Files broken into chunks (typically 64 MB) • Chunks triplicated across three machines for safety • See SOSP^03 paper at hTp://labs.google.com/papers/gfs.html
Rep
licas
Master GFS Master
GFS Master Client
Client
C1 C0 C0
C3 C3 C4
C1
C5
C3
C4
WISE Keynote 68 2010-12-13
![Page 69: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/69.jpg)
WISE Keynote � 69 2010-12-13
![Page 70: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/70.jpg)
Win in Scale�
• Google Translate • Voice • Trend PredicAon
– An example benefits society�
WISE Keynote � 70 2010-12-13
![Page 71: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/71.jpg)
H1N1 United NaAon Report �
WISE Keynote � 71 2010-12-13
![Page 72: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/72.jpg)
Concluding Remarks • Search + Social • Increasing quanAty and complexity of data demands scalable
soluAons • Have parallelized key subrouAnes for mining massive data sets
– Spectral Clustering [ECML 08] – Frequent Itemset Mining [ACM RS 08] – PLSA [KDD 08] – LDA [WWW 09, AAIM 09] – UserRank [Google TR 09] – Support Vector Machines [NIPS 07]
• Launched Google Q&A (Confucius) in 60+ countries • Relevant papers
– hTp://infolab.stanford.edu/~echang/ • Open Source PSVM, PLDA
– hTp://code.google.com/p/psvm/ – hTp://code.google.com/p/plda/
WISE Keynote 72 2010-12-13
![Page 73: Integra(ng)Social)with)Search) - Stanford Universityinfolab.stanford.edu/~echang/Wise2010-Keynote.pdf · Related&Papers • AdHeat’(Social’Ads): – AdHeat:&An&Influence9based&Diffusion&Model&for&PropagaAng&Hints&to&Match&Ads,&](https://reader035.vdocuments.site/reader035/viewer/2022070916/5fb6f837ce87433be65eb2bf/html5/thumbnails/73.jpg)
WISE Keynote 73
Models of InnovaAon • Ivory tower
• Only consider theory but not applicaAon
• Build it and they will come • ScienAsts drives product development
• “Research for sale” • Research funded by:
– product groups or customers
• Research & development as equals • Research “sells” innovaAon; • Product “requests” innovaAon
• Google-‐style innovaAon
R
R
D
D
R
D R
R=D
D?
2010-12-13