基于语料库的有标并列短语语体适应性考察

Research on the Marked Coordinate Phrases′ Flexibility of Language Style Based on Language Database

本文通过一个多语体小型语料库,考察现代汉语有标并列短语的语体适应性,发现不同语体在有标并列短语的出现频率、标记类型、并列项类型及多项式选择等方面都有显著差异,得出一些具体结论:有标并列短语语体分布的优选序列为"口述体<对话体<操作体<新闻体<法规体",不同标记类型有所差别;并列项的优选序列为"数量短语/代词<形容词<异类<动词<名词",但口述体、对话体中异类词、代词更为优选;多项式优选序列为"口述体<对话体<新闻体<操作体/法规体"。本文的结论将为并列短语的自然语言识别和处理提供参考性帮助。

his article mainly researches the flexibility of language style about marked coordinate phrases in modern Chinese language through a small language database of multiple styles. The research discovers the significant differences of different lan- guage styles on the aspects of frequency of occurrences, marked styles ~, styles of coordinate items and multiple choice. It comes to some concrete conclusions. The prior choice for marked coordinate phrases of language style distribution is ＂oral 〈 conversa- tion aspect 〈 operating 〈 journalese 〈 law and regulation＂. There are differences between different marked types. The prior choice for coordinate items is ＂numeral phrase/pronoun 〈 adjective 〈 alien words 〈 verb 〈 noun＂, but alien word and pronoun are prior to others in oral aspect and conversation. The prior choice for multiple items is ＂oral 〈 conversation 〈 journalese 〈 operat- ing 〈 law and regulation aspect＂. These conclusions will be of referential value to the natural language discrimination and treat- ment of coordinate phrases.