0.25
0.5
0.75
1.25
1.5
1.75
2
Detecting Erroneous Identity Links on the Web using Network Metrics
Published on Nov 22, 20182875 Views
Although best practices for publishing Linked Data encourage the re-use of existing IRIs, multiple names are often used to denote the same thing. Whenever multiple names are used, owl:sameAs statement
Related categories
Chapter list
Detecting erroneous identity links in the web of data00:00
Motivation00:08
'identity' links in the lod00:41
"The sameas problem" (1/2)00:52
"The sameas problem" (2/2)02:04
How can we detect erroneous sameas links?02:47
Requirements03:48
Approach & Experiments04:30
Overall idea04:38
Dataset05:29
1. Extract the explicit identity statements05:37
2. Partition to equality sets05:52
'Barack Obama' equality set06:09
3. Detect the community structure in each eq set06:32
4. Assign error degrees07:19
Communities - 'Barack Obama'07:59
Evaluation08:46
Error degree distribution of 556m owl:sameas08:49
Objectives09:04
Finding the threshold (1/3)09:28
Finding the threshold (2/3)10:26
Evaluation / Accuracy (1/2)12:09
Finding the threshold (3/3)12:54
Evaluation / Accuracy (2/2)13:09
Evaluation / Recall13:55
Who is messing up the lod? - 114:49
Who is messing up the lod? - 215:12
Who is messing up the lod? - 315:43
Conclusion & perspectives16:49
summary16:50
Perspectives17:59
Thank you!18:22