Although the combination of OWL and Horn rules results in the creation of a highly expressive language, i.e. SWRL, there are still many occasions where this language fails to accurately represent knowledge of our world. In particular, SWRL fails at representing vague and imprecise knowledge and information. Such type of information is apparent in many applications like multimedia processing and retrieval, information fusion, etc. In this paper, we propose f-SWRL, a fuzzy extension to SWRL to include fuzzy assertions (such as ‘Mary is tall in the degree of 0.9’) and fuzzy rules (such as ‘being healthy is more important than being rich to determine if one is happy’).
Jeff Z. Pan, Giorgos B. Stamou, Vassilis Tzouvaras