SWRL-F – A Fuzzy Logic Extension of the

Semantic Web Rule Language

Tomasz Wiktor Wlodarczyktomasz.w.wlodarczyk@uis.no

2

Plan

• Goals• Design Choices• SWRL-F ontology• SWRL-F rules• Conclusions

3

Goals

• To be able to express and utilize fuzzy assertions in SWRL using a standard ontology editor

4

Design Choices - General

• Express fuzzy knowledge using OWL classes• Fuzzy inference in SWRL-F is limited to the rules

only• Follow fuzzy control systems scheme: fuzzification,

inference, defuzzification

5

Design Choices - Particular

• Represent fuzzy assertions - object property• Reuse already existing reasoner – FuzzyJess• FuzzyJess is a superset of Jess • Automatically compatibility with existing built-ins

6

SWRL-F ontology

7

SWRL-F ontology

8

SWRL-F ontology

9

SWRL-F ontology

10

SWRL-F ontology

11

SWRL-F ontology

12

SWRL-F rules

• Curve(?c) ∧

hasCurveDelayDuration(?c, ?d) ∧

fuzzymatch(?d, ShortCurveDelayDuration)

∧

hasRisk(?c, ?r) →

fuzzymatch(?r,

LowRisk)

13

SWRL-F rules

14

Conclusions

• extension to SWRL that allows constructing fuzzy rules in SWRL based on language variables described using OWL-based ontology

• general design is based on fuzzy control systems• SWRL-F allows extending any SW application with

FL capabilities basing on Protégé editor and modified SWRLJessTab

• can be embedded into other applications using Protege API and JDBC

SWRL-F – A Fuzzy Logic Extension of the

Semantic Web Rule Language

Tomasz Wiktor Wlodarczyktomasz.w.wlodarczyk@uis.no