Week3

From INLS 490

Jump to: navigation, search

Week 3

Contents

[edit] Video

Six degrees of separation by Surfca13

[edit] Week 3

[edit] Feedback and Critique

Questions and critique of the readings. What interesting themes or ideas emerged? What challenged you? What did you buy or not buy?

[edit] Network dynamics

[edit] Networks 101

  • Nodes, edges and paths
    • Nodes: The objects to be connected in a network
    • Edges: The ties that bind nodes in a network
    • Paths: Our means of transversal between places in a network
  • Concepts of distance and betweenness
    • Network analytic models compute the distance between nodes calculated in path transversals
    • Betweenness: The connenctive quality of a node measured by its place in the graph
    • Distance: The shortest path length between a node in a graph

[edit] Graph Models

  • Erdos-Renyi Random Model
    • Erdos-Renyi assume that edges are non-preferential, ensuring a random distribution of connections in a graph.
    • How does this model fail in real life?
  • Watts-Strogats "Small Worlds" Model
    • The Small World model predicts that edges display clustering characteristics that would account for some non-randomness (Bell Curve).
  • Scale-Free Network
    • Barabasi et. al. predict that edges display preferential attachment, creating a power-law distribution of attachments in the network. This accounts for hubs and authorities in a network.

[edit] Centrality

  • What is centrality and why is it important?
    • Erdos/Kevin Bacon and Six Degrees of Separation
    • A-list bloggers. How does the "scale-free" blogosphere benefit A-list bloggers?
    • How would we test these models in Facebook or other Social Network?

[edit] Information flow in networks

  • How do we establish edges in networks
    • On the web
    • On social networks
    • In real world social networks
  • Structural constraints in networks
    • What about the web, or Facebook, or any other information medium enforces "network law"
    • How do we "break" network law (on the web, in SNS, etc?)

[edit] Critique of network study

  • What does a single-viewpoint study of a network topology tell us?
  • What are some of the problems with applying network-theoretic approaches to our real-world networks?
  • What does it mean if we value all friends alike?

[edit] Tools for analyzing social networks

  • Pajek, UCINet, Jung

[edit] Questions

  • Relationship between voting and a blog's ranking - do power laws promote poor content?
  • To get a holistic view of a network, how many means of connection should we account? Since different communication tools have different qualities, how can we factor these in?
  • Does the world get smaller because of networks?
  • Bloggers aren't charitable, are links votes?
  • How does the virtual space we occupy affect our information choice capacity?
  • How do social networks collapse the borders between individuals?
  • Can we nail down exactly the interaction between networks and blog popularity?
  • How do we negotiate authority in the blogosphere?
  • The role of marketing in popularity (using your networks)
  • What other web-type systems display small-world/scale-free characteristics?
  • How would we develop a SNS test of the 6 Degrees hypothesis
  • Qualitative and intervening factors in graph growth - how do things like age and friendship quality affect network growth?
  • What effect do spikes have in network growth? (Digg, Slashdot)?
  • The taxonomy of friendship:
    • Electronic friend connection
    • Electronic message
    • Instant messaging
    • Written note
    • Telephone call
    • Face-to-face, no travel required
    • Face-to-face, travel required
  • Blogroll as the A-list.


[edit] Presentations

[edit] Quick Links