Study on Semantic Transparency of Chinese Compounds Based on Word Embedding

摘要

Based on word embeddings, this paper proposes an automatic calculation method for the semantic transparency of Chinese compounds and realizes the calculation of the semantic transparency of a large number of compounds. According to the experiment results, we find that the grammatical structure of Chinese compounds and the semantic relationship between compounds and its morphemes are vital factors affecting the semantic transparency of compounds. We prove that the semantic transparency of compounds with parallel grammatical structure is higher than that of compounds with other grammatical structures. There are two important semantic relationships between compounds with high semantic transparency and its internal morphemes: AB≈A≈B; AB≈A+B, by analyzing compounds with higher transparency.