Large-Scale Graph Mining

Large-Scale Graph Mining

Vincent Leroy

FREQUENT SUBGRAPH MINING (FSM)

? Graphs represent complex data

? Chemical compounds, proteins ? Social networks ? Knowledge bases (ontologies) ? Frequent Subgraph Mining

10M entities, 120M facts

? Discover regularities in the structure of a graph

? Properties and interactions (citations graph, organization structure)

? Privacy (social networks)

570M entities, 18B facts (2012)

? Link prediction (recommender systems, linked data)

2

FSM: CHEMISTRY

Caffeine molecule C8H10N4O2

3

FSM: CHEMISTRY

Caffeine molecule C8H10N4O2

3

FSM: KNOWLEDGE BASE

Jim Halpert works at works at

Dwight Schrute

lives in

Dunder Mifflin

located in

Scranton

lives in

Don Draper works at works at

Peggy Olson

Sterling located in Cooper

lives in

New York

4

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download