Computationally efficient sup-t transitive closure for sparse fuzzy binary relations (Q812624)

For the use of fuzzy relations in knowledge representation it is important to determine their sup-t-transitive closures. The authors consider fuzzy relations over finite universes of discourse, which thus are representable by square matrices, and assume that the representing matrices for their fuzzy relations are sparse matrices. For this particular case the paper offers two algorithms to determine transitive closures, and it discusses, theoretically and with computational tests, their complexity and, shows that they have lower computational complexity than the standard approach for the general case.