分词语料库中四字格的切分和识别研究-语言学及应用语言学专业论文.docx

PAGE PAGE 5 中文摘要 四字格的能产性和派生性极强，利用四字格模式创造出的新词数量在现代汉 语词汇中仍然呈上升趋势，因此对四字格的研究不能仅仅局限于文献和理论。本 文将研究的目光投向分词语料库中数量众多的四字格，并针对分词语料库中四字 格易被切碎的特点，进行了四字格抽取、四字格切分不一致比较、四字格识别等 一系列工作。 本文首先对分词语料库中的四字分词单位进行提取和筛选，得到四字格抽取 结果；并利用四字格抽取结果，进行了分词语料库内部和分词语料库之间的四字 格切分不一致比较工作。在四字格识别研究中，通过引入 crf 统计模型，并将四 字格切分不一致结果作为模型训练语料，本文在有词性标注分词语料库中进行了 四字格的识别研究。利用 crf 模型识别得到的四字格结果，统计四字格用字、词 性信息并观察内部结构特点总结规则，本文在无词性标注分词语料库中也进行了 四字格的识别研究。识别结果表明，在有词性标注分词语料库和无词性标注分词 语料库中，四字格的识别精度都能达到 90%左右。 关键字：分词语料库，四字格抽取，切分不一致，CRF Abstract The productive and derivative of four-character idioms are extremely high, the use of four-character pattern to create new words in the vocabulary of modern Chinese still on the rise, so the works on four-character idioms can not be limited to research and theoretical literature. This article will look into the eyes of the large number of four-character idioms in word-segmented corpora, and works for extraction, segmented comparison, recognition and a series of work of four-character idioms for the easily shred characteristics of four-character idioms in word-segmented corpora. This article first works on the fourth sub-word units in word-segmented corpora for extraction and screening, in order to take the results as the four-character idioms extraction; and works on the segmented comparison of four-character idioms both in single segmented corpora and between different segmented corpora by using the results of four-character idioms extraction . In the works of four-character idioms recognition, through the introduction of crf statistical model, and take the results of segmented comparison of four-character idioms as a training corpora of the model, in this article we develop the research of the recognition of four-character idioms in POS-tagged corpora. By using the results of the recognition of four-character idioms based on Crf model, and statisticsing the words, POS information, interal features of four-character idioms to summarize rules, in this article we also develop the research of