Measuring Semantic Entropy
I. Dan Melamed · 1997
Semantic entropy is a measure of semantic ambiguity and uninformativeness. It is a graded lexical feature which may play a role anywhere lexical semantics plays a role. This paper presents a method for measuring semantic entropy using translational distributions of words in parallel text corpora. The measurement method is well-defined for all words, including function words, and even for punctuation. 1 Introduction Semantic entropy is a measure of semantic ambiguity and uninformativeness. This paper presents a method for measuring semantic entropy using translational distributions of words in parallel text corpora. The measurement method is well-defined for all words, including function words, and even for punctuation. The hypothesis behind the measurement method is that semantically heavy words are more likely to have unique counterparts in other languages, so they tend to be translated more consistently than semantically lighter words. The consistency with which words are translat...