Introduction SNA
Introduction SNA
• Introduction
• Definitions
• Basic Concepts
Course Learning Outcomes (CLO)
Social
Networking
Wiki
Forum
Social Media: Many-to-Many
Various forms of Social Media
7
Characteristics of Social Media
9
Characteristics of Social Media
• Participation
• social media encourages contributions and feedback from everyone who is interested. It blurs the
line between media and audience.
• Openness
• most social media services are open to feedback and participation. They encourage voting,
comments and the sharing of information. There are rarely any barriers to accessing and making
use of content – password-protected content is frowned on.
• Conversation
• whereas traditional media is about “broadcast” (content transmitted or distributed to an audience)
social media is better seen as a two-way conversation.
• Community
• social media allows communities to form quickly and communicate effectively. Communities share
common interests, such as a love of photography, a political issue or a favorite TV show.
• Connectedness
• Most kinds of social media thrive on their connectedness, making use of links to other sites,
resources and people.
Introduction
Why Networks?
15
Networks: Rich data
Networks of the Real-world [1]
• Information networks:
• World Wide Web
• Citation networks
• Blog networks
• Social Networks:
• Organizational networks
• Communication networks Cita%on networks
• Collaboration networks
• Technological Network:
• Power grid
• Airline, road, river networks
• Telephone networks
• Internet
• Autonomous systems
Networks of the Real-world [2]
• Biological networks:
• Metabolic networks
• Food webs
• Neural networks
• Gene regulatory networks
• Language networks:
• Semantic networks
• Software networks:
• Call graphs
Sample of an online social network Protein interaction network
Sweden’s economic network of interlocked corporations
Three models of epidemic spread in human contact networks.
Definitions
Social Network Analysis
Definitions
• Actor/Node/Vertex
• Edge/Relation/Tie
• Path
• Network/Graph
Actor/Node/Vertex
Undirected
Directed
Graph
CTI-3A3
Applied Social Network Analysis
Big Data
Mining Social Network Data
• Traditional obstacle:
• Large-scale
• Realistic
• Completely mapped
• Now: large on-line systems leave detailed records of social activity
• On-line communities: MySpace, Facebook, LiveJournal
• Email, blogging, ecommerce, instant messaging
• On-line publications repositories, arXiv, MedLine
Networks: A Matter of Scale
• Web as a graph
• Google PageRank
• How to estimate webpage importance from the structure of the web-
graph?
• Routing in peer-to-peer networks:
• BitTorrent, ML-donkey, Kazaa, Gnutella
• Can we find a file in a network without a central server?
Applications [2]
• Marketing and advertising:
• How to define influence?
• How to find influencers?
• Who to give free products to
• so that we create a network effect?
• Diffusion of information and epidemics:
• How to trace information as it spreads?
• How to efficiently detect epidemics
and information outbreaks?
Applications [3]
• Friend/link prediction:
• How to predict/suggest friends in
networks?
• Trust and distrust:
• How to predict who are your
friends/foes? Who to trust?
• Community detection:
• How to find clusters and small
communities in social networks
Level of Social Network Analysis
• Nodal/Actor level
• focuses on nodal level attributes and phenomena
• Dyadic level
• focuses on the pairs of nodes
• Triadic level
• focuses on triplets of nodes
• N-adic/Subset level
• focuses on sub-graphs of N nodes
• Network/Group level
• focuses on the whole graph and network level Typically a cross-level analysis, combining all of these levels
phenomena
Challenges
• Scale
• This work considers “extreme-scale” graphs – billion+ vertices and up to
trillion+ edges
• Processing these graphs requires at least hundreds to thousands of
compute nodes or tens of thousands of cores
• Graph analytic algorithms are generally memory-bound instead of
compute-bound; in the distributed space, this results in a ratio of
communication versus computation that increases with core/node count
Complexity
58
Social Computing Tasks
• Social Computing: a young and vibrant field
• Conferences: KDD, WSDM, WWW, ICML, AAAI/IJCAI,
SocialCom, etc.
• Tasks
– Network Modeling
– Centrality Analysis and Influence Modeling Our Focus
– Community Detection
– Sentiment Analysis, Classification and Recommendation
– Privacy, Spam and Security
59
Network Modeling
• Large Networks demonstrate statistical patterns:
• Small-world effect (e.g., 6 degrees of separation)
• Power-law distribution (a.k.a. scale-free distribution)
• Community structure (high clustering coefficient)
• Model the network dynamics
• Reproducing large-scale networks
• Examples: random graph, preferential attachment process, Watts and Strogatz
model
• Simulation to understand network properties
• Thomas Shelling’s famous simulation: What could cause the segregation of white and black
people
• Network robustness under attack
Centrality Analysis and Influence Modeling
• Centrality Analysis:
• Identify the most important actors or edges
• E.g. PageRank in Google
• Various other criteria
• Influence modeling:
• How is information diffused?
• How does one influence each other?
• Related Problems
• Viral marketing: word-of-mouth effect
• Influence maximization
61
Community Detection
• A community is a set of nodes between which the interactions are (relatively)
frequent
– A.k.a., group, cluster, cohesive subgroups, modules
62
Classification and Recommendation
• Common in social media applications
• Tag suggestion, Product/Friend/Group Recommendation
Link predic-on
Network-Based Classification
63
Privacy, Spam and Security
• Privacy is a big concern in social media
• Facebook, Google buzz often appear in debates about privacy
• NetFlix Prize Sequel cancelled due to privacy concern
• Simple annoymization does not necessarily protect privacy
• Spam blog (splog), spam comments, Fake identity, etc., all requires new techniques
• As private information is involved, a secure and trustable system is critical
• Need to achieve a balance between sharing and privacy
64
Sen$ment Analysis in
Social Network Analysis
Example of Opinion
66
Source: Bing Liu (2011) , “Web Data Mining: Exploring Hyperlinks, Contents, and Usage Data,” Springer, 2nd Edition,
Sentiment Analysis vs Subjectivity Analysis
Sentiment Subjectivity
Analysis Analysis
Positive
Subjec5ve
Negative
Neutral Objective
67
Levels of Sentiment Analysis
Sentiment Analysis
Vishal Kharde and Sheetal Sonawane (2016), "Sentiment Analysis of Twitter Data: A Survey of Techniques,"
International Journal of Computer Applications, Vol 139, No. 11, 2016. pp.5-15
Subjectivity
Classification
Polarity
Determination
Approaches
Sentiment Vagueness Machine Learning based
Classification resoluCon in
opinionated
Review text Lexicon based
Sentiment
Usefulness
Analysis Measurement
Multi- & Cross-
Hybrid
Lingual SC
approaches
Opinion Spam Cross-domain
DetecCon SC
Source: Kumar Ravi and Vadlamani Ravi (2015), "A survey on opinion mining
Application and sentiment analysis: tasks, approaches and applications." Knowledge-Based
Systems, 89, pp.14-46.
69
Sentiment Classification Techniques
Support Vector
Decision Tree Machine (SVM)
Classifiers
Supervised Neural Network
Machine Learning Linear (NN)
Classifiers Deep Learning
Learning
(DL)
Approach Rule-based
Classifiers Naïve Bayes
(NB)
SenFment Unsupervised Probabilistic
Analysis Learning Bayesian
Classifiers Network (BN)
Dictionary- Maximum
Lexicon- based Entropy (ME)
Approach
based
Approach StaFsFcal
Corpus-based
Source: Jesus Serrano-Guerrero, Jose A. Olivas, Francisco P. Romero, and Enrique Herrera-Viedma (2015),
"SenFment analysis: A review and comparaFve analysis of web services," InformaFon Sciences, 311, pp. 18-38.
Approach SemanFc
Features and Metrics in Social Network
• Sentiment Analysis is one of the most used methods adopted to analyze data collected
through online social networks à can investigate the opinions and attitudes expressed
online by means of natural language processing tools
• Problems: it does not allow one to consider data within the online network in which they
have been collected.
• Social Network Analysis, through a quantitative-relational approach, makes it possible to
consider data as “networked” (ie, considering existing connections and links between
users).
How to Integrate SNA in SA?
• The properties of the linkage between individuals on online social networks are critical to
an understanding of the process of social influence through them à represented by the
sociological construct of tie strength that represents the strength of the dyadic
interpersonal relationships in the context of social networks
• Comment between 2 people is useful for predicting edge signs and may be used to fit a
conventional sentiment model. A purely edge feature–based sentiment model cannot
account for the network structure since it reasons about edges as independent of each
other.
Conclusions
• Network:
• Scale
• Structure
• Information Diffusion
• Challenges
• Heterogeneity
• Evolution
• Evaluation
References
• Community Detection and Mining in Social Media. Lei Tang and Huan Liu,
Morgan & Claypool, September, 2010.
• Newman, ME, The Structure and Function of Complex Networks, SIAM
2003
• Watts, DJ; Strogatz, S H. 1998.
Collective dynamics of 'small-world' networks, NATURE 393(668).
• Barabasi, AL. Network Science Network Science
CII-6A3
Pengantar Komputasi Sosial
Thank You