The ‘K’ in ‘Semantic Web’ Stands for ‘Knowledge’

19th June 2018

Wouter Beek (

Simplifying assumptions of KR research

[1/6] Quantity Assumption

“There is only little data”

The LOD Laundromat Stack

See Chapter 2: Scaling Data Cleaning to the Web

[2/6] Quality Assumption

“Statements are well-defined and interpretable”

Most data cannot be read (e.g. Freebase 14%).

See Chapter 4: Scaling Data Quality to the Web

[3/6] Accessibility Assumption

“Data can be easily (re)used”

SPARQL endpoint availability in 2016.

See Chapter 3&5: Scaling Querying/Metadata to the Web

[4/6] Homogeneity Assumption

“Results over few datasets can be generalized to all datasets”

Datasets used in ISWC 2014 evaluations.

See Chapter 6: Scaling Evaluations to the Web

[5/6] Context-independence Assumption

“Meaning is independent of context”

Things can be (considered) the same in some, but not all contexts.

See Chapter 7: Scaling identity to the Web

[6/6] Declarativeness Assumption

“Meaning is captured by formal semantics”

Graph A

                abox:store tbox:sells abox:tent.
                abox:tent  tbox:costs "¥150,000".
                abox:tent  rdf:type   abox:Product.
Graph B

                fy:aHup   pe:ko9sap_ fy:jufn12.
                fy:jufn12 pe:oao9_   "Ufou".
                fy:jufn12 rdf:type   fyufnt:tmffqt.

See Chapter 8: Scaling Naming to the Web

Thank you for your attention!