520 likes | 551 Views
Learn about various network types and their properties, examples of networks, and the interdisciplinary approaches to study them. This course covers topics like social, information, technological, economic, and biological networks, and their influences on behavior and systems dynamics.
E N D
MIS 644 Social Network Analysis Chapter 1 Introduction 2017/2018 Spring
Outline Networks Why to study networks Examples of networks Properties of networks Outline of the course
Networks • Collection of objects in which some pairs are connected by links • Verticies or nodes actors in SNA • Links adges
Why to study networks • Individual parts or components linked by some way • Examples • Internet: collection of computers linked by communication links • social networks: people by friendship relations • Individual components • How a computer works • How a humen being feals or acts • Connections or interactions • Communication protocoles TCP/IP • Dynamics of human friendship – social psychology
Third approach (1) • Patterns of connections • Network - simplified reprsentation abstraction • components - nodes • connections - links • Patterns of interactions • affects behavior of systems • e.g.: connection patterns – Internet • affect – routers, efficiency of tranfering data • e.g.:: in social networks • how people learn, form opinions or gether news • spread of disease • Structure - behavior
Third approach (2) • Verious disiplines • Mathematics, computer science, statistics, sociology, economics, ... • analysing, modeling and understanding networks • Many tools – simple network representation • verticies and edges • e.g.: • Which is the best connected vertex • Length of path from one vertex to another • Network models – how links are formed • Processes on networks - Predictions about: • The way trafic flow over Internet • The way diseases spread through a community
Third approach (3) • Not always works • Depens on • What the system is and does • Specific questions to be answered • Networks are • general and powerful • way of representing • patterns of connection and interraction of components
Examples • A classification by N-N • Technological networks • Social and economic networks: • Information networks • Biological netwoks
Technological Networks • Internet • Electrical power grids • Phone networks • Transport networks: • Roads, airline connection between airports, rail roads and subway networks
The Internet • Vertecies – computers, rooters • Edges – physical data connections • optical fibers or telephone lines • Man made, carefully engineered • Its structiure is not known exactly • Many different groups, little centrilized control • See Figure 1.1 of N-N • Why study the network structure of the Internet? • How to choose a route? • Is the shortest route the fastest? • How to avoid buttlenecks in the trafic flow? • What happens when a vertex or edge fails? • Where to add new capacity? • Development of new standards or protocoles
Information Networks • Logical connections • www • E-mail • Phone calls • Nodes: telephone number • Directed or weihted links: number of calls from A to B • Trust network • Link: one accepts other public key • Citations • Social media
Web • Network of information stored in Web pages • Vertcies – web pages • Edges – hyperlinks - software constract • Content structure – link structure • Useful page gets many links • Web – directed network • Run in a specific direction: from one page to another • Dificulty: many pages are dynamically generated
Information Networks • www: links among Web pages • How pages are related • How groups of communities • Which pages are more prominent or important • Search engines - Google • Network structure - evaluate • quality and relevance of web pages • Not number of links it receives • A page is prominent if receives link from prominent pages
Social Networks • Collection of social ties among friends • A network of people or group of people such as firms • Vertecies - people or groups (organizations) • Edges – connections of some kind • friendship or acquaintances , working relations, sexual • business relations • Sociology • Increase in complexity due to • travel, technological advences, global communication, digital interaction • Exampe: Actor network • Wayne Zachary’s karate club - small 34 members • Figure 1.2 of N-N • Facebook or MS Instant messaging – large data
the pattern of emailcommunication among 436 employees of Hewlett Packard Research Lab
Economic Netwoks • Special types of social networks • Vertecies- firms, countries, industries • Edges – trading relationships between companies or countries- • Global manifacturing operations – network of suppliers • Media companies – network of advertizers • Spread of local breakdown to • Cascading failures or financial crisis
Biological Networks • Logical abstract • Interractions between • genes and/or proteins • molecules in the cells metabolic pathways • Interractions physical but links not • Possibility of inerraction molecules • Food web: Ecological, predetor-pray networks • Vertecies – species • Directed Links – predation of one by onother • Physical: • nervous system, neurons in the brain • blood vessels
How do we reason about networks? • Empirical: Study network data to find organizational principles • How do we measure and quantify networks? • Mathematical models: graph theory and statistical models • Models allow us to understand behaviors and distinguish surprising from expected phenomena • Algorithms for analyzing graphs • Hard computational challenges
Properties of Networks • Structure • Behavior • Dynamics • On networks – processes on .- given structure. • Of networks – evolution/coevolution
Structure • How to investigate? • Visualization – small or medium size networks with relatively sparse links • Measures or metrics • e.g.: centrality: • Quantifies how important vertices are • Degree of a node: number of edges attached to a node • Hubs
Some Topics • Structure of netowks • Small World Phenomena • Community Detection
Structure • A language for structural features of networks • Complxities • Parts that are more or less densely interconnected • Central cores • Participants can be more central or peripherial
Hubs: vertices with unusually high degree • Observed in many networks • Algorithms for • analyzing and understaning • network data • Measurements of network properties
Small world effect • Milgram experiment • 6 degrees of seperation • Internet • Computer are few hops away
Clusters or Communities in Networks • Social networks – break down into subcommunities • Knit friends or acquantions • Business relationships of companies • Clustered sets • Community detection techniques
the network structure of political blogs prior to the 2004 U.S. Presidentialelection
Behavior and Dynamics • Structure – starting point • Connectedness at the level of structure • Who is linked to whom • At the level of behavior • Individual’s actions – consequenses • Outcomes of anyone
Processes on Networks • Failure and resilience • Diffusion epidemics • Search
Dynamics on and of networks • coevolution of networks • structure – proceses on networks
Network Dynamics – Structural Effects • How people influence each other • Taking into acount structure of the network • The underlaying mechanisms • İnformation and • population level • Local level - network • İn the structure links friends • Aligning behavior with immediate friends rather then population as a whole
Search • The way people explore social contact for information • Effectiveness of performing these tasks – structure of networks
Outline of the course • Prerequisites for MIS 644 • Work load
Prerequisits for MIS 644 • elementary knowledge of linear algebra • matrices, tranpose, inverse of a matrix, eigen values and vectors • calculus • elementary probability and statistics • basic knowledge of algorithms • preferentially experinece with a programming language • some experience in data processing and manipulation
Work Load • Homeworks • Project • mid reports • final report • presentation • Paper presentations • mostly applied papers • network model of a online social network • Software evaluation • examination and evaluation of a software • (programable softare not software witha GUI only)
Course Structure • Basic tools • Network structure • Network models • Statistical analysis of networks – ERGMs • Processes on networks
Basic Tools and Stgructure • basic theoretical tools used to • describe and analyze networks, • graph theory: • measures and metrics for quantifying networkstructure • patterns revealed in real-world networks • mathematics and metrics • commnity detection
Network Models • Mathematical modelsof networks • Random graphs and extnsions • Small-world model • Network growth models • preferntial atacments and extensions • exponential random graph models • game theretic models
Detailed Course Outlines • Introduction to networks • Course overview • Basic Concepts • Graph theory • Verious types of networks • Metrics • path, components, • Centrality metrics • degree, betweenness, closeness, • eigen value, • Clique, clusters • Paths, components, degree distribution, clustering, degree correlations (assortativity) • Metrics for weighted spatial networks • Properties of real networks
Properties of real networks • Shortest paths and small-world property • Scale-free property and heavy-tailed degree distributions • Clustering coeficients • Hierarchy - Modularity • Network motifs • Network models • Random networks (Poisson and with a given degree distribution) • Watts-Strogatz model • Preferential attachment and its variants • Optimization-based network formation models • Exponential random graphs
Community detection methods • Strength of weak ties • Graph partitioning - Focus on spectral partitioning algorithm • Modularity maximization methods • Spectral partitioning • Hierarchical clustering - divisive and agglomerative methods • Overlapping communities, dynamic communities, • Statistical properties of community structures • Statistical analysis of network data • Network sampling methods • Bias in network sampling • Network inference based on cross-correlations
Dynamics of networks • Percolation and network resilience • Random and non-random removel of vertices • Network epidemics • SI, SIS, SIR models • Network immunization • Identification of major spreaders • Social networks and influence • Information and behavior spreading - measurements and models • Influence maximization • Outbreak detection