Unsupervised Learning of Data Linking Configuration thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Unsupervised Learning of Data Linking Configuration

Published on Jul 04, 20123633 Views

As commonly accepted identifiers for data instances in semantic datasets (such as ISBN codes or DOI identifiers) are often not available, discovering links between overlapping datasets on the Web is g

Related categories

Chapter list

Unsupervised Learning of Link Discovery Configuration00:00
Link discovery problem00:02
Instance matching01:02
How to avoid manual configuration?02:15
What is a good decision rule?03:04
Evaluation measures03:32
Reference datasets04:20
Assumptions05:10
Fitness criterion06:13
Fitness function: unsupervised case07:25
Approach – genetic algorithm09:06
Encoding decision rules10:13
Mutation10:32
Crossover10:59
Implementation11:17
Fitness function behaviour12:13
Genetic algorithm parameters13:40
OAEI 2010: Person/Restaurant15:14
OAEI 2011: New York Times data16:10
Reducing computations17:02
Reducing computations: sampling17:59
Alternative indirect evidence18:55
Fitness function comparison20:05
Summary20:54
On-going & future work21:20
Questions?22:58