CS224W: Mining and learning graphs1

来源：Jure Leskovec, Stanford University

Why Networks?

Networks are a general language for describing complex systems.
There are complex systems all around us:

• Society is a collection of 7+ billion individuals
• Communication systems link electronic devices
• Information and knowledge are organized and linked
• Interactions between thousands od genes/proteins regulate life
• Our thoughts are hidden in the connections between billions of neurons in our brain

Networks!

Behind many systems there is an intricate wiring diagram, a network, that defines the interactions between the components.
We will never be able to model predict these systems unless we understand the networks behind them!

Many Types of aData are Networks

• Social networks
• Information networks
• Economic networks
• Internet
• Communication networks
• Patient networks
• Hierachies of cell systems
• Gene co-expression networks
• Disease pathways
• Cell-cell similarity networks

Why Networks？Why Now?

• Universal language for describing complex data
  Networks from science, nature, and technology are similar than one would expect
• Shared vocabulary between fields
  Computer Science, Social Science, Physics, Economics, Statistics, Biology
• Data availability & computational challenges
  Web/mobile, bio, health, and medical
• Impact
  Social networking, Social medis, Drug design
  
  

Ways to Analyze Networks

• Predict the type/color of a given node
  Node classification
• Predict whether two nodes sre linked
  Link prediction
• Identify densely linked clusters of nodes
  Community detection
• Measure similarity of two nodes/networks
  Network similarity
  
  

This reveals two important themes of this class:

• We must understand how netwoork structure affects the robustness of a system
• Develop quantitative tools to access the interplay between network structure and the dynamical processes on the networks, and their impact on failures
• We will learn that in reality failures follow reproducible laws, that can be quantified and even predicted using the tools of networks
  

Application

Misinformation

• Q: Is a given Wikipedia article a hoax?
• Real articles link more coherently:
  
  Link Prediction
  Content recommendation is link prediction
  
  Polarization on Twitter
  
  Predicting Virality

Product Adoption

Function Prediction
Classifying the function of proteins in the interactome.

Reasoning about Networks

• Patterns and statistical properties of network data
• Design principles and models
• Algorithms and predictive models to answer questions and predictions

How do we mine networks?

• Empirically: Study network data to find organization 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

What do we study in networks?

Structure and evolution

• What is the structure of a network?
• Why and how did it come to have such structure?
  Processes and dynamics
• Networks provide a "skeleton"for spreading of information, behavior, diseases
• How do information and diseases spread?