论文部分内容阅读
本研究以上海交通大学科技英语语料库(JDEST)为基础,运用计量语言学的研究方法,以协同语言学为理论框架,对英语复合词的生成趋向进行了实证研究。研究重点探讨了英语复合词的生成数量与词干的长度、词干在语料库中出现的频数和词干义项数三种属性之间的依存关系。研究结果证实了协同语言学中的相关理论假设,表明在一定长度的语料中,词干出现频数越高,生成复合词的种类越多;词干义项数越多,生成复合词的种类越多;但词干越长,生成复合词的种类越少。研究同时表明幂率模型y=axb能够准确描述词干长度和词干频数与复合词生成数量之间的关系,但就词干义项数而言,幂率模型y=a+bx2具有更高的拟合度。
Based on the JDEST (Sci-Tech English Corpus) in Shanghai Jiaotong University, this study uses the method of econometrics and the collaborative linguistics as the theoretical framework to make an empirical study on the generation of English compound words. The research focuses on the relationship between the number of generated English compound words and the length of the stem, the frequency of occurrence of the stem in the corpus, and the number of stemmed items. The results of the study confirm the relevant theoretical assumptions in co-linguistics, indicating that in a certain length of corpus, the higher the frequency of stemming occurs, the more types of compound words are generated; the more the stemming terms are, the more types of compound words are generated; The longer the stem, the less the kinds of compound words. The study also shows that the power rate model y = axb can accurately describe the relationship between stem length and stem frequency and the number of compound words. However, for the number of stemmed terms, the power model y = a + bx2 has a higher Coincidence.