Evaluating the premises and results of four metaphor identification systems

摘要: This study first examines the implicit and explicit premises of four systems for identifying metaphoric utterances from unannotated input text. All are then evaluated on a common data set in order to see which most successful. The goal is if these can find metaphors corpus that mostly non-metaphoric without over-identifying literal humorous as metaphors. Three distributional semantic systems, including source-target mapping method [1-4]; word abstractness measurement [5], [6, 7]; similarity [8, 9]. fourth knowledge-based system uses domain interaction based SUMO ontology [10, 11], implementing hypothesis metaphor product interactions among all concepts represented an utterance [12, 13].