Upload
seung-hyun-seo
View
344
Download
0
Embed Size (px)
Citation preview
KAIST Software Graduate School
Software Eco system 지도교수 김진형 1
과목 : Software Eco system학번 : 20113920
이름 : 서승현
Software Eco system 지도교수 김진형 2
1. History of Facebook ……………………………………………………………………………
3
2. Business Model ……………………………………………………………………………………
5
2.1 Ads …………………………………………………………………………………………………
6
2.2 Third parties …………………………………………………………………………………
7
3. Facebook Technology for Business ………………………………………………………
8
3.1 Social Graph …………………………………………………………………………………
9
3.2 News Feed(Edge Rank) …………………………………………………………………
10
3.3 Social Plugin …………………………………………………………………………………
11
4. VS Google …………………………………………………………………………………………
12
4.1 Why Google? …………………………………………………………………………………
13
4.2 Open Social ……………………………………………………………………………………
14
4.3 Page Rank ………………………………………………………………………………………
15
5. Future of Facebook ………………………………………………………………………………
17
5.1 Network Effect ………………………………………………………………………………
18
5.2 Link Predict ……………………………………………………………………………………
20
5.3 Open Graph API ……………………………………………………………………………
23
5.4 Timeline …………………………………………………………………………………………
25
6. Conclusion ……… …………………………………………………………………………………
26
6.1 Evolve into Interest Graph ………………………………………………………………
27
6.2 Consider with Side Effect ………………………………………………………………
28
References
Software Eco system 지도교수 김진형 3
Software Eco system 지도교수 김진형 4
2004
2006
2005
2008
2007
2010
2009
2011
users
year
500 mi
1. History of Facebook
Software Eco system 지도교수 김진형 5
6Software Eco system 지도교수 김진형
2. Business Model 2.1 Ads
7Software Eco system 지도교수 김진형
Zynga Wants your relationship Friends information
Foursquare Wants your route path Hot Spot
Analyze your pattern
2. Business Model 2.2 Third parties
Software Eco system 지도교수 김진형 8
9Software Eco system 지도교수 김진형
Facebook’s Connect make Relationship Group Community Echo System Private Web
how they're related
Make Private Web Use Facebook Social Graph Service use Facebook’s Information Share user’s log between Facebook and
Web
3. Facebook Technology for Business
3.1 Social Graph
10Software Eco system 지도교수 김진형
Newsfeed Latest Posts and Popular Posts Popular Post is default configuration We don’t know where configuration exists Facebook has ‘The Will’ that make private newsfeed
Marketing It is important for Facebook to show ‘Popular Posts’ We saw information to newsfeed than ‘like page’
Support Social Graph Affinity, Weight , Time Decay Analyze relationship
3. Facebook Technology for Business
3.2 Edge Rank
11Software Eco system 지도교수 김진형
3. Facebook Technology for Business
3.3 Social Plug-in
Software Eco system 지도교수 김진형 12
13Software Eco system 지도교수 김진형
Google have Infra and Money already Gmail Searching Engine Android, Motorola
Open Social Platform of Platform It is mine, and yours is mine too
Also launch ‘Google+’ Is Linked in Gmail address And Google docs… and so on
4. VS Google 4.1 Why Google?
14Software Eco system 지도교수 김진형
State of Union Common API Project Facebook vs Non-Facebook ? App Store vs KT,SK,LG ? Non-Facebook = Google Plus + Open Social = Google
4. VS Google 4.2 Open Social
15Software Eco system 지도교수 김진형
4. VS Google 4.3 Page Rank
16Software Eco system 지도교수 김진형
Page Rank equals Google Business Advertisement Searching Trend Prediction Page rank is not unique method but it is important
Collective Intelligence You are not in, but you’re in by Searching Detecting influenza epidemics using search engine query
data SNS consist of active participation Google realize relationship from personal action
4. VS Google 4.3 Page Rank
Software Eco system 지도교수 김진형 17
18Software Eco system 지도교수 김진형
5. Future of Facebook 5.1 Network Effect
19Software Eco system 지도교수 김진형
Key Elements People Identity Type of Relationships Relationships Identity
Facebook Power is Network Size The value is dependent on the number of others using. Expect positive feedback more useful the more users join like telephone Ads Target is Relationship and Identity
Consequently, User is Money
5. Future of Facebook 5.1 Network Effect
20Software Eco system 지도교수 김진형
5. Future of Facebook 5.2 Link Prediction
21Software Eco system 지도교수 김진형
Method of making more network effect Prevent to leaving platform Lower 30% can’t be adapted Making Group Six degrees of separation Person maybe you know
But, want too much undergo cumbersome feel In fact, don’t like that
5. Future of Facebook 5.2 Link Prediction
22Software Eco system 지도교수 김진형
5. Future of Facebook 5.2 Link Prediction
23Software Eco system 지도교수 김진형
Moto is Private web MS-docs, Pandora, Slide-Share Real-time Crawl user’s interest information Private Service Social Plugin is part of Open Graph strategy
Facebook Know your Interest and hobby
Not just Service, be the Platform
But, Alert Privacy
5. Future of Facebook 5.3 Open Graph
24Software Eco system 지도교수 김진형
5. Future of Facebook 5.3 Open Graph
25Software Eco system 지도교수 김진형
Term was used in twitter already Allocate contents as time passed Contents Meaningless as time decay
Facebook Say “We give you Memory” Photo, video, etc.. Making diary
5. Future of Facebook 5.4 Timeline
Software Eco system 지도교수 김진형 26
27Software Eco system 지도교수 김진형
6. Conclusion6.1 Evolve into Interest Graph
6. Conclusion6.1 Evolve into Interest Graph
28Software Eco system 지도교수 김진형
Interest Prediction
Interest Discovery
Difficult to know what is real interest
29Software Eco system 지도교수 김진형
6. Conclusion6.2 Consider
with Side Effect
30Software Eco system 지도교수 김진형
We know Facebook exert ‘Positive Effect’ Wider relationship Interaction Opportunity Information Power of ‘the public’
But, Do you really want ‘cool relationship’? Make friends easily, but lose easily too Many friends, but lonely
Manipulate public opinion
6. Conclusion6.2 Consider
with Side Effect
Software Eco system 지도교수 김진형 31
32Software Eco system 지도교수 김진형
[1] Facebook’s Edge Rank : How to Make Sure You’re in the News Feed(Buddy Media)
[2] Open Editing Algorithm: A Collaborative News Promotion Algorithm based on Users’ Voting History(2009, GSCT KAIST)
[3] Multi-Relational Link Prediction in Heterogeneous Information Networks(Darcy Davis, Ryan Lichtenwalter, Nitesh V. Chawla ,University of Notre Dame)
[4] Identifying and evaluating community structure in complex networks(2009, Karsten Steinhaeuser, Nitesh V. Chawla , University of Notre Dame)
[5] The Anatomy of a Large-Scale Hypertextual Web Search Engine(Sergey Brin and Lawrence Page ,Stanford University)
[6] The PageRank Citation Ranking: Bringing Order to the Web(Stanford University) [7] The Google Pagerank Algorithm and How It Works(2002, Ian Rogers ,
sirgroane.net) [8]The Link Prediction Problem for Social Networks( 2004, David Liben-Nowell, Jon
Kleinberg ,MIT, Cornell University) [9] IT 정책연구 시리즈 제 16 호 (2011 , @ 한국정보화진흥원 ) [10] The Interest Graph and What Marketers Should Know About It (Stefan Herbert,
Published December 9, 2011) [11] http://en.wikipedia.org/wiki/Positive_feedback
Reference
Software Eco system 지도교수 김진형 33