A Fuzzy Semantics for the Resource Description Framework

A Minimal Deductive System for General Fuzzy RDF

“…The most general work so far and closest to our work is [14], to which respect we provide additionally a more general semantics, correctness and completeness and complexity results, add the notion of top-k answers of the union of conjunctive queries in which answers may be scored by means of a scoring function, and show how to compute the top-k answers. Another related work is [21], which allows to annotate triples with truth values taken from a finite partial order, while we rely on [0, 1] instead 1 .…”

“…Essentially, [21] does not provide a truth combination function to propagate the truth in such inferences, while we consider a t-norm instead. Additionally, as for [14], the top-k retrieval problem for the union of conjunctive queries is not addressed.…”

“…We now present fuzzy RDF in its general form, by extending [12,13,14]. To do so and to make the paper self-contained, we first recall basic notions of mathematical fuzzy logic [9].…”

“…However, under the classical semantics, RDF cannot represent vague information and, to this purpose, some Fuzzy RDF variants have been proposed [12,13,14,21,22]: essentially they allow to state that a triple is true to some degree, e.g., (tom, likes, tomato) is true to degree at least 0.9.…”

“…The advantage is that, (i) (unlike [12,13,14,22]) we abstract from the underlying XML representation; (ii) the semantics is quite general, i.e. is based on a t-norm [9]; (iii) we get a clear insight of the supported inference mechanism; and (iv) we concentrate on the main ingredients of RDF from a reasoning point of view.…”