Decision Lists for Lexical Ambiguity Resolution: Application to Accent Restoration in Spanish and French

摘要

This paper presents a statistical decision procedure for lexical ambiguityresolution. The algorithm exploits both local syntactic patterns and moredistant collocational evidence, generating an efficient, effective, and highlyperspicuous recipe for resolving a given ambiguity. By identifying andutilizing only the single best disambiguating evidence in a target context, thealgorithm avoids the problematic complex modeling of statistical dependencies.Although directly applicable to a wide class of ambiguities, the algorithm isdescribed and evaluated in a realistic case study, the problem of restoringmissing accents in Spanish and French text.