The emerging field of computational social science (CSS) is devoted to the pursuit of interdisciplinary social science research from an information processing perspective, through the medium of advanced computing and information technologies.
1. PREM SANKAR C
Research Scholar
Dept of Futures Studies, University of Kerala
Computational Social Science
It's is all about data and connections .....
KNM Govt Arts and Science College
3. What is Computational Social
Science?
Cyberculture Studies Digital Dataset + Computational Methods
4. 1. Definition : CSS
The increasing integration of technology into our lives
has created unprecedented volumes of data on
society’s everyday behaviour. Such data opens up
exciting new opportunities to work towards a
quantitative understanding of our complex social
systems, within the realms of a new discipline known
as Computational Social Science.
(Conte at al. 2012)
6. Study of how humans
organize and interact in our
modern society....
Discovery of new patterns and their quantitative description in socio-economic
systems.
Interface between computer science and the traditional social
sciences
7. Era of Electronic Word of Mouth Communication
Decentralized control and Self-organization - Group behavior emerges from the
local interactions of individuals (Gender Studies / Collective Actions)
#Metoo
Social relationships are hidden
ICT - Huge Volume of Data
Internet /Social media/ Data Mining
=>Computational Social Science /
Digital Humanities
Interview
Questionnaire
Survey
10. Computational Social Science (CSS)
CSS is a research discipline at the interface between
computer science and the traditional social sciences.
It uses computational methods to analyze and model
social phenomena, social structures, and collective
behavior.
11. Computer Science + Social Science
Computer Science
Study anything
Methods driven
Large data
Prediction
Social Science
Study social things
Question driven
Small data
Explanation
Feel the Difference
15. Digital Trace Data: Examples
1) Social media sites
2) Web search queries
3) Blogs and internet forums
4) Call detail records from mobile phones
5) Sensor data
Online networks
Call logs
Instant messaging
There has been a fundamental shift in the opportunities in data
collection about humans.
Data-driven social science
16. Social Science in Digital Era !!!
Digital Datasets can provide information that can profoundly
shape our understanding of social phenomena.
Convert these data sets into computational models
Simulation of complex social interactions or Computational
approaches with data to estimate models of social
phenomenon.
The main computational approaches to the social sciences are
1) Social network analysis
2) Social mining
3) Social complexity modeling
4) Social simulation models.
Data is the new oil
17. From Data to Theory?
Sociological Theory
Hypothesis / Prediction
Web Data
We need theory, because we want to explain social issues.
18. CSS Methodologies
Case study analysis
Controlled experiments
Computational modeling
Integrative data analysis / natural experiments
20. What is SNA?
Social network analysis (SNA) is a collection of techniques,
tools, and methods by which one can map and analyze the
connections across individuals or groups or institutions.
Network analysis allows us to examine how the structure of
networks.
Why SNA?
SNA is multidisciplinary and deals with
• Influencing groups (public health, propaganda, viral marketing)
• Increasing engagement with stakeholders (Management/Recommendation)
• Cool algorithms/ Analytics (math, computer science)
• Study of social behaviour (sociology, cognitive science)
• Organizational behavior (leadership, management)
21. Actors (Nodes/ Vertices)
Actors –are the smallest unit of a network
Persons
Organizations
Countries
Companies
Animals
Words
Web pages
Families
22. Relations (Link/Edge/Tie/Arc)
Kinship
mother of, wife of
Other role-based
boss of, teacher of
friend of
Cognitive/perceptual
knows
aware of what they know
Affective
likes, trusts
Interactions
give advice, talks to, fights
with, lends money to
sex / drugs with
Affiliations
belong to same clubs /
companies
is physically near
Two Actors are connected
by a social relationship.
23. Type of Relations
Relations can be
Undirected
Directed
Weighted
Weight can be
Strength
Rank
Frequency
Probability
24. “Think Link”
Social network
Collaboration network
Terrorist networks
Economic networks
Family Networks
Organization networks
Sports Networks
Co-author Networks
A BIs related to
Patterns are
left behind
See your
interconnected
world
26. Network Measures…
1- Centrality Measures
Identifying people who are well positioned to influence
the network or to move information around.
Network Science
28. Identifying Key People
Who are the people who are best positioned to move information through
the network?
People who live in the intersection of social worlds are at higher risk of
having good ideas. –Ron Burt
32. It’s a small world, after all ...
“six degrees of separation—everyone in the
world is connected to everyone else through a
chain of, at most, six mutual connections.
In Milgram’s 1967 “small world experiment”,
individuals were asked to reach a particular
target individual by passing a message along a
chain of acquaintances. For successful chains,
the average # of intermediaries needed was 6.
Director: Fred Schepisi
Writers: John Guare (play), John
Guare (screenplay)
Stars: Will Smith, Stockard
Channing, Donald Sutherland
1993
33. How connected are you to everyone else in
the world?
The Facebook average is 3.57.
Source - Facebook Data Science Report
Estimated average degrees of separation between all people on Facebook.
The average person is connected to every other person by an average of 3.57
steps. The majority of people have an average between 3 and 4 steps
34. Three Degrees of Influence
your friends’ friends’ friends
In Book Connected by
Nicholas A. Christakis
and James H. Fowler.
35. The Strength of Weak Ties
Weak ties are more likely to be bridges to outside networks than strong ties
(emotionally close friends and family) .
Weak ties provide access to information and resources beyond those
available in their own social circle.
New information and ideas.
37. The Tipping Model
How to identify the small “seed” group of people who can spread a
message across an entire network for ViralMarketing .
38. Super-spreaders / Hubs
Image -206 SARS patients diagnosed in Singapore were traced to four super-spreaders.
Patient Zero, the physician from China, who brought the disease to Singapore.
നനിപ്പ ബബാധയുണണ്ടെനന്ന് കണണ്ടെതനിയ 18 പപരനിൽ 16 പപരരും മരണമടഞ്ഞു ഈ പതനിനബാറു പപരരും
ആദദരും മരണമടഞ്ഞ സബാബനിത്തുമബായനി ബന്ധമള്ളവരബായനിരന
Gamifying Epidemic - http://vax.herokuapp.com/game
How do you control the behaviour of a network?
39. Eco Chambers -How Close Are You Really?
If you are part of a group of close friends or relations, you are less able
to make strong links outside this group.
40. Homophily
Homophily is the tendency to connect with people with similar
characteristics (status, beliefs, etc.)
41. Social Networks Expands
Idea - ‘the friends of my friends are my friends’:
The probability of three people being friends with each other in a
social network
The familiar strangers we see every day on the bus and in the supermarket
43. A) Network Dynamics & Social Behavior
How do revolutions emerge without anyone expecting
them?
How did social norms about same sex marriage change
more rapidly than anyone anticipated?
Why do some social innovations take off with relative
ease, while others struggle for years without spreading?
What are the forces that control the process of social
evolution –from the fashions that we wear, to our beliefs
about religious tolerance..
We don't need emperors or even centralized institutions
to get these kinds of social phenomenon
44. Diffusion of behaviors on Facebook
Posts, Share, Nomination(mention in post),Volunteer (in
comments)
Selfi Culture
45. Diffusion of #Hoaxs #Misinformation, #FakeNews
Each piece of Digital misinformation contributes to the
shaping of our opinions.
Fake news make more money than real one
Clickbait sites manufacture hoaxes to make money from
ads, while news sites publish and spread rumors and
conspiracy theories to influence public opinion.
Eco champers exists - Demonetization fans spreads that
the Amartya Sen endorsed it, others spread that he
opposed it. He did both.
Trust in social networks -give higher priority to more
reliable sources.
46. B) Social Navigation - #Collective Actions
#Kerala Floods
#Sabarimala
# Metoo
#AyodhyaCase
Idea # using information from nearest neighbors
(follow the known crowd)
* Social information reuse.
* Trust in Social Networks –
1) Charity Activities
47. 47
Emotion on Facebook
Classify semantic content of status updates using LIWC
Emotion: fraction of posts with positive/negative words
Coviello et al., PLoS ONE 2014, “Detecting Emotional Contagion in Massive Social Networks”
Slides provided by Lorenzo Coviello. Thanks! Later partially modified.
48. 48
Twitter for Migration Studies
Use streaming API filter for geo-tagged tweets from
OECD countries
Pick 3,000 users per country, get their tweets
Estimate out-migration and oversample “static”
countries
Get data for ~500K users
After activity thresholding left with ~15K
E. Zagheni, K. Garimella, I. Weber, B. State: Inferring International
and Internal Migration Patterns from Twitter Data. WWW’14
52. Blumenstock, Joshua, Gabriel Cadamuro, and Robert On. “Predicting poverty and wealth
from mobile phone metadata.” Science 350, no. 6264 (2015): 1073-1076
1) Mobile Phone Data
53. Jost, John T., Pablo Barbera, Richard Bonneau, Melanie Langer, Megan Metzger, Jonathan Nagler,
Joanna Sterling, and Joshua A. Tucker. “How social media facilitates political protest:
Information, motivation, and social networks.” Political psychology 39 (2018): 85-118.
2) Twitter & Political Movements
54. Garcia, David, and Bernard Rime. “Collective emotions and social resilience in the digital
traces after a terrorist attack.” Psychological science (2019): 0956797619831964.
3) Twitter & Social Emotions
55. 4) Facebook and Gender Gaps
Figure: Gender gaps in internet use computed using data from Facebook
(online
model) available at www.digitalgendergaps.org
Fatehkia, Masoomali, Ridhi Kashyap, and Ingmar Weber. “Using Facebook ad data
to track the global digital gender gap.” World Development 107 (2018): 189-209.
56. Ginsberg, Jeremy, Matthew H. Mohebbi, Rajan S. Patel, Lynnette Brammer, Mark S. Smolinski, and
Larry Brilliant. “Detecting influenza epidemics using search engine query data.” Nature 457, no.
7232 (2009): 1012.
5) Google Search - Flu Detection
57. 6) UBER Data - Taxi Meters in NYC
Farber, Henry S. “Why you can’t find a taxi in the rain and other labor supply lessons from cab
drivers.” The Quarterly Journal of Economics 130, no. 4 (2015): 1975-2026.
58. 7) Chat bots - Anti-racism bots
Munger, Kevin. 2017. “Tweetment Effects on the Tweeted: Experimentally
Reducing Racist Harassment.” Political Behavior.
60. Project 1 # Board Interlock Network
• Interlocking director refers to the situation in which
the same person shares positions on the boards of
more than one company.
61. Results..
Paper : C. Prem Sankar, K. Asokan, K. Satheesh Kumar*, Exploratory Social
Network Analysis of Affiliation Networks of Indian Listed Companies Social
Networks, Social Networks, 43, 113–120 2015.
Existence of small world structure in the Indian corporate sector.
Power elight: 2.25% of the director population account for 65.5% of the total market
capitalisation.
62. C. Prem Sankar, K. Asokan, K. Satheesh Kumar*, Exploratory Social Network Analysis of
Affiliation Networks of Indian Listed Companies Social Networks, Social Networks, 43,
113–120 2015.
63. Project 2 # Cochin Metro Network
Paper : Analysis of road network of the buffer area of Kochi Metro
rail using tools of social network analysis
64. Project 3 # MF Investment Network
Paper : Trust based Stock Recommendation System -a
Social Network Analysis Approach
65. Project 5 # Twitter Networks
#kissoflove
Zonin Alessandro ,Digital
Marketing Manager in
IBM Italy
Zonin Alessandro ,Digital
Marketing Manager in
IBM Italy
A link between
User A and User B
exists if one has
mentioned the
other in a tweet
A link between User and Hashtag exists if user has mentioned a hashtag
in their tweet
66. Project 4 # Linguistic Network
Independence day
Speeaches
• Each word is a node and
their co-occurrence is
encoded as the edge
between them
Paper : Forecasting and Modeling Long Term Policy Change Using Semantic
Linguistic Networks :Case Study in Indian Context
67. Project 7 #MakeInIndia
Paper : Social Network Visualization on How ‘MakeInIndia’ Made Vibes Among
Various Sectors – A Topic Modeling Based Approach
69. Part1 # Influence Maximization Algorithms
In this work, We proposed a new node ranking method to measure
the influence of influencer’s in a social context by analyzing the social
interaction.
70. Part 2 # Agent based Simulation
In complex social systems, when traditional approaches fails to capture the dynamics,
ABMS has been reported to be successful in capturing emergent phenomena resulting
from the interactions of individual agents.
Not Exact Prediction - It provides Explanation and Experimentation
72. Research Designs
1) Digital platforms & society: testing theories
2) Combining digital trace data with conventional data
sources like surveys
3) Apps for data generation and extraction
4) Combining digital traces with experiments (e.g. bots)
5) Ethics / Privacy Issues
6) Political Polarisation
7) Group Dynamics
73. 73
Research Categories
Relational (qualitative)
• Strength of ties
• Accessibility
• Likeability/”fun”
• Reputation
• Expected reciprocity?
• Competing unit?
• Dependence
• Trust
Structural
(quantitative)
• Size
• Density
• Diversity
• Structural
Holes
• Isolates/Clique
s
• Centrality
• Betweeness
• Closeness
Individual (qualitative)
• Personality (e.g., Big
5, self-monitoring)
• Emotional
intelligence
• Intentionality
• Past experience
• Sentimental analyis
75. Profiling by Association
Imagine that
You’re on Facebook
Your profile is empty
But you have friends in Facebook
All PhD students
All based in Kerala (Locality / School)
Mostly from one city or one college
Are you “anonymous”?
Really hard to anonymize network information!