Sign inSign up
Multi-Label Relational Neighbor Classification using Social Context Features thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Multi-Label Relational Neighbor Classification using Social Context Features

Published on Sep 27, 20137424 Views

Xi Wang

Networked data, extracted from social media, web pages, and bibliographic databases, can contain entities of multiple classes, interconnected through different types of links. In this paper, we focus

KDD 2013 - Chicago

Related categories

Data MiningKnowledge Extraction

Chapter list

Multi-label Relational Neighbor Classification using Social Context Features00:00
Motivation00:19
Problem Formulation01:41
Classification in Networked Data02:06
Contribution02:54
Relational Neighbor Classifier - 103:33
Relational Neighbor Classifier - 204:17
Apply RN in Multi-relational Network04:43
Edge-Based Social Feature Extraction06:04
Cluster edges using K-Means07:10
Edge-Clustering Visualization07:54
Proposed Method: SCRN08:32
SCRN09:16
SCRN Overview10:01
SCRN Visualization10:41
Datasets - 111:18
Datasets - 211:47
Datasets - 312:04
Comparative Methods12:16
Experiment Setting - 112:54
Experiment Setting - 213:54
Results (Micro-F1)14:31
Results (Macro-F1)15:18
Results (Hamming Loss) - 115:28
Results (Hamming Loss) - 215:52
Results (Hamming Loss) - 316:13
Conclusion16:54
Reference17:52
Thank you17:55