Solving symbolic constraints (approximately)

Date: 29/06/2021
Time: 09:00 - 09:50

Location: online via Zoom

ABSTRACT: Symbolic constraint solving is ubiquitous in many areas of mathematics and computer science. Unification, matching, anti-unification, disunification, and ordering constraints are some prominent examples that play an important role in automated reasoning, term rewriting, declarative programming, and their applications. In this talk we present our recent results (obtained jointly with C. Pau) about solving unification and anti-unification constraints in theories where equality is replaced by its quantitative approximation. Such problems arise in working with imprecise information, requiring results "within tolerance". Our algorithms solve unification and anti-unification problems modulo crisp and fuzzy tolerance (i.e., reflexive and symmetric) relations, where mismatches between symbol names and arities are permitted.