Semantic Similarity of Inverse Morpheme Words Based on Word Embedding

Inverse morpheme words are compound words that have the same morphemes but are arranged in the opposite order. The majority of related works on the subject have focused on a narrow investigation of dictionary definitions, with few studies based on large-scale corpora. Based on the People's Daily corpus (1946-2017), we add and delete words from a base list and then obtained a word list consisting of 668 pairs of inverse morpheme words. Furthermore, we also calculated cosine similarity by using word embedding based on the distributed representation and discovered that 76% of inverse morpheme words have a cosine similarity of 0.4 or higher, and that word formation, part-of-speech, and frequency all have an impact on semantic similarity.