D1
D2
D3

SocialLens

Searching & Browsing Communities by Content and Interaction

Motivation

Social Network Data

Social data is noisy and in large scale, hence it is hard for human to find the useful/desired information

Our Goal

Provide a tool (like lens) to zoom in/out the whole network so as to discover the useful signals from the big social graph

SocialLens: Searching & Browsing Communities by Content and Interaction

Input

Output

A social graph G=(V,D,E,L)

  • V: users
  • D: documents published by users
  • E: friendship links between two users
  • L: diffusion links among two documents

Example:

  • A set of communities:
    • Community membership: who is in each community?
    • Community profile
      • Content profile: what is each community about?
      • Diffusion profile: how do communities interact with each other in a topic?

Dataset selection:



What is each community?

How does each community interact with others?

How to make use of community profile?

Keywords: