SLIDE 1
node2vec: Scalable Feature Learning for Networks
Aditya Grover, Jure Leskovec
Farzaneh Heidari
Outline
- word2vec (Background)
- Random Walk (Background)
- node2vec
- Evaluation Results
- Deficiencies
node2vec: Scalable Feature Learning for Networks Aditya Grover, - - PowerPoint PPT Presentation
node2vec: Scalable Feature Learning for Networks Aditya Grover, - - PowerPoint PPT Presentation
node2vec: Scalable Feature Learning for Networks Aditya Grover, Jure Leskovec Farzaneh Heidari Outline word2vec (Background) Random Walk (Background) node2vec Evaluation Results Deficiencies 3 4 Random word2vec
SLIDE 2
SLIDE 3
3
SLIDE 4
4
SLIDE 5
5
Random Walk word2vec node2vec
SLIDE 6
word2vec
SLIDE 7
word2vec’s backbone
SLIDE 8
Window in Graph
SLIDE 9
Random Walk
Stochastic Process Path of random steps
SLIDE 10
10
Feature Learning in Graphs
Goal: Learn features for a set of objects Feature learning in graphs: § Given: § Learn a function:
§ Not task specific: Just given a graph, learn f. Can use the features for any downstream task!
55 Jure Leskovec, Stanford
SLIDE 11
11
Feature Learning in Graphs
Goal: Learn features for a set of objects Feature learning in graphs: § Given: § Learn a function:
§ Not task specific: Just given a graph, learn f. Can use the features for any downstream task!
55 Jure Leskovec, Stanford
Unsupervised Feature Learning
§ Intuition: Find a mapping of nodes to d-dimensions that preserves some sort of node similarity § Idea: Learn node embedding such that nearby nodes are close together § Given a node u, how do we define nearby nodes?
§ !
" # … neighbourhood of u obtained by
sampling strategy S
56 Jure Leskovec, Stanford
SLIDE 12
12
Feature Learning in Graphs
Goal: Learn features for a set of objects Feature learning in graphs: § Given: § Learn a function:
§ Not task specific: Just given a graph, learn f. Can use the features for any downstream task!
55 Jure Leskovec, Stanford
How to determine !" #
Two classic search strategies to define a neighborhood of a given node:
for !" # = 3
u s3 s2
s1
s4 s8 s9 s6 s7 s5
BFS DFS
58 Jure Leskovec, Stanford
SLIDE 13
13
BFS vs. DFS
Structural vs. Homophilic equivalence
BFS: Micro-view of neighbourhood
u
DFS: Macro-view of neighbourhood
59 Jure Leskovec, Stanford
SLIDE 14
14
Feature Learning in Graphs
Goal: Learn features for a set of objects Feature learning in graphs: § Given: § Learn a function:
§ Not task specific: Just given a graph, learn f. Can use the features for any downstream task!
55 Jure Leskovec, Stanford
BFS vs. DFS
Structural vs. Homophilic equivalence
BFS-based:
Structural equivalence (structural roles)
DFS-based:
Homophily (network communities)
60 Jure Leskovec, Stanford
SLIDE 15
15
Interpolating BFS and DFS
§ Biased random walk procedure, that given a node # samples !" #
v
α=1 α=1/q α=1/q α=1/p
x2 x3 t x1
The walk just traversed (),+) and aims to make a next step.
61 Jure Leskovec, Stanford
SLIDE 16
16
Feature Learning in Graphs
Goal: Learn features for a set of objects Feature learning in graphs: § Given: § Learn a function:
§ Not task specific: Just given a graph, learn f. Can use the features for any downstream task!
55 Jure Leskovec, Stanford
Multilabel Classification
§ Spectral embedding § DeepWalk [B. Perozzi et al., KDD ‘14] § LINE [J. Tang et al.. WWW ‘15]
Algorithm Dataset BlogCatalog PPI Wikipedia Spectral Clustering 0.0405 0.0681 0.0395 DeepWalk 0.2110 0.1768 0.1274 LINE 0.0784 0.1447 0.1164 node2vec 0.2581 0.1791 0.1552 node2vec settings (p,q) 0.25, 0.25 4, 1 4, 0.5 Gain of node2vec [%] 22.3 1.3 21.8
62 Jure Leskovec, Stanford
SLIDE 17
Trade-offs
task-specific heuristics inefficient usage of statistics
SLIDE 18
18