摘要：This research establishes Database of Morpheme Meaning Annotations for Compounds, and statistically analyses the using frequency of morphemes and their meanings in the database. The research shows that the using frequency of “zi” (子) and “er” (儿) is the highest, and their meanings with the highest using frequency are applied when the morphemes are affixes. Among various meanings of one morpheme, only one meaning would be applied with high frequency, with simpler and specific implications. By overviewing all morphemes, it is observed that those with higher using frequency do not necessarily have meanings with high frequency of use. By analyzing internal structures of compounds, this paper finds that noun-modification compounds are the most, followed with verb-object compounds and affixation compounds. The semantic logic relations between morphemes and compound words can be: word definitions can be the abstraction of objects signified by morpheme meanings, or of senses signified by morpheme meanings, or of spatial meanings signified by morpheme meanings.
关键词：compound words; morpheme meaning; database; word formation
