|
Tác giả CN
| 于涛. | |
Nhan đề
| 英语动词型式自动识别研究 /于涛； 梁茂成. | |
Mô tả vật lý
| tr.366-378. | |
Tóm tắt
| This paper aims to automatically extract English verb patterns based on similarity measure and clustering of concordances. The results show that 1) different verbs, the number of concordances and the heterogeneous level of concordances all influence the precision of the clustering of concordances and that the selection of the number of groups based on the internal validity of KMeans algorithm is much more flexible and open, which yields better results and higher precision; 2) average precision indices of automatic pattern extraction are 90.99% and 95.91% respectively in the two clusters of concordances involved, higher than the average precision achieved in previous studies(81%), with parentheses and special sentence structures as the main causes influencing the precision and recall of the automatic extraction of English verb patterns. It is believed that the automatic pattern identification and extraction model proposed in this study is quite feasible and applicable. | |
Tóm tắt
| 本文采用相似度分析的方法,在语料库索引行自动聚类的基础上实现英语动词型式的自动识别与提取。研究发现:第一,检索词的不同、索引行的数量及索引行间的差异程度三者共同影响索引行聚类的准确率;基于聚类内部效度确定索引行分组数量的方法更为灵活、开放,结果也更为可靠,准确率更高。第二,两次聚类中英语动词型式自动识别平均准确率分别达到90.99%和95.91%,均高于前人研究中81%的平均准确率。此外,通过对错误分组索引行的分析发现,导致型式自动识别错误的主要因素为插入语成分及特殊句式。本文所提出的型式自动识别与提取模型不仅具有准确率高和灵活变通的特点,而且具有广泛的适用性。 | |
Thuật ngữ không kiểm soát
| 型式识别. | |
Thuật ngữ không kiểm soát
| 型式语法. | |
Thuật ngữ không kiểm soát
| 聚类. | |
Tác giả(bs) CN
| 梁茂成. | |
Nguồn trích
| 外语教学与研究- vol.49, no.3 (May 2017) |
| |
000
| 00000nab a2200000 a 4500 |
|---|
| 001 | 50709 |
|---|
| 002 | 2 |
|---|
| 004 | 61492 |
|---|
| 008 | 171012s ch| a 000 0 chi d |
|---|
| 009 | 1 0 |
|---|
| 022 | |a10000429 |
|---|
| 039 | |y20171012092352|zhuett |
|---|
| 041 | 0 |achi |
|---|
| 044 | |ach |
|---|
| 100 | 0 |a于涛. |
|---|
| 245 | 10|a英语动词型式自动识别研究 /|c于涛； 梁茂成. |
|---|
| 300 | |atr.366-378. |
|---|
| 520 | |aThis paper aims to automatically extract English verb patterns based on similarity measure and clustering of concordances. The results show that 1) different verbs, the number of concordances and the heterogeneous level of concordances all influence the precision of the clustering of concordances and that the selection of the number of groups based on the internal validity of KMeans algorithm is much more flexible and open, which yields better results and higher precision; 2) average precision indices of automatic pattern extraction are 90.99% and 95.91% respectively in the two clusters of concordances involved, higher than the average precision achieved in previous studies(81%), with parentheses and special sentence structures as the main causes influencing the precision and recall of the automatic extraction of English verb patterns. It is believed that the automatic pattern identification and extraction model proposed in this study is quite feasible and applicable. |
|---|
| 520 | |a本文采用相似度分析的方法,在语料库索引行自动聚类的基础上实现英语动词型式的自动识别与提取。研究发现:第一,检索词的不同、索引行的数量及索引行间的差异程度三者共同影响索引行聚类的准确率;基于聚类内部效度确定索引行分组数量的方法更为灵活、开放,结果也更为可靠,准确率更高。第二,两次聚类中英语动词型式自动识别平均准确率分别达到90.99%和95.91%,均高于前人研究中81%的平均准确率。此外,通过对错误分组索引行的分析发现,导致型式自动识别错误的主要因素为插入语成分及特殊句式。本文所提出的型式自动识别与提取模型不仅具有准确率高和灵活变通的特点,而且具有广泛的适用性。 |
|---|
| 653 | 0 |a型式识别. |
|---|
| 653 | 0 |a型式语法. |
|---|
| 653 | 0 |a聚类. |
|---|
| 700 | 0 |a梁茂成. |
|---|
| 773 | |t外语教学与研究|gvol.49, no.3 (May 2017) |
|---|
| 890 | |a0|b0|c0|d0 |
|---|
| |
Không có liên kết tài liệu số nào
|
|
|
|