en-es
en-fr
en-sl
en
0.25
0.5
0.75
1.25
1.5
1.75
2
From Relational to Semantic Data Mining
Published on Nov 28, 20171207 Views
Relational Data Mining (RDM) addresses the task of inducing models or patterns from multi-relational data. One of the established approaches to RDM is propositionalization, characterized by transformi
Related categories
Chapter list
Relational and Semantic Data Mining00:00
Slovenia – Ljubljana (capital)00:27
Jožef Stefan Institute, Ljubljana, Slovenia00:39
Department of Knowledge Technologies01:27
Department of Knowledge Technologies01:33
Talk overview01:38
Background: Data mining 01:49
Predictive data mining: Learning a decision tree classifier02:17
Predictive data mining: Learning classification rules02:19
First Generation Data Mining02:22
Second Generation Data Mining03:06
Subgroup Discovery03:35
Example SD application: High CHD Risk Group Detection04:19
Example SD application: Functional genomics05:16
SD algorithms in the Orange DM Platform06:37
Second Generation Data Mining Platforms07:15
Second Generation Data Mining Platforms07:51
Relational data mining08:37
Relational data mining 09:49
Relational and semantic data mining10:03
Using domain ontologies in RDM 11:21
Semantic Data Mining12:26
Our early work: Semantic subgroup discovery13:08
Propositionalization approach to SDM 14:23
Propositionalization approach to SDM - Relational subgroup discovery (RSD)15:09
Semantic subgroup discovery with RSD15:35
Step 2: RSD feature construction17:43
Step 3: RSD Propositionalization18:39
Step 4: RSD rule construction with CN2-SD19:18
Semantic Data Mining in Orange4WS20:00
Semantic subgroup discovery with SEGS20:35
Semantic subgroup discovery with SEGS - 121:27
BioMine graph exploration engine (Toivonnen et al.)21:36
SegMine: Complex semantic data mining methodology22:40
SegMine implementation in Orange4WS platform24:24
SDM-SEGS: Generalizing SEGS25:06
SDM-Aleph: Generalizing Aleph26:22
Hedwig26:56
Third Generation Data Mining Platform: ClowdFlows27:59
ClowdFlows platform29:43
ClowdFlows platform - 130:51
ClowdFlows platform - 231:07
Talk overview 32:32
Challenge addressed in current work33:50
Challenge addressed in current work - 134:56
Methodology: Step 135:56
Methodology: Step 236:21
Methodology: Step 336:31
The methodological framework36:43
Experimental finding38:32
NetSDM algorithm outline39:22
Results39:37
Summary and conclusion: SDM in context40:21
Summary and conclusions41:12
Summary and conclusion: Future work46:29
Paradigm shift in Semantic Data Mining: Mining Linked Open 47:21
Acknowledgements48:10
Department of Knowledge Technologies49:05