论文部分内容阅读
蒙古文信息处理工作在应用开发方面都已从字、词处理阶段进入到句处理阶段。我们知道,字处理阶段的工作基本不涉足语义问题,而词处理阶段,尤其到了句处理阶段,对词语、短语和句子的理解、生成等都需要多方面的语义知识,包括词语的语义分类、语义搭配特征和语义角色等方面的详细描述。蒙古语当代语义学研究成果在很大程度上偏向于语义概念的抽象解释,而忽略了语义外层的具体表现形式。因此也无法满足信息处理的实际需求。蒙古语的语义知识的多数,是在句法结构中得到解释,而现阶段蒙古文信息处理的句法分析却是亟待解决的问题。以蒙古语的高频词和真实语料为主要素材,进行了语义分类、语义场划分、义素分析、配价及格语法等方面的探索性研究,取得了一些认识,也掌握了一些切实有效的语义分析方法。
Mongolian information processing work in the application development has been from word, word processing stage into the sentence processing stage. We know that the basic stages of word processing do not involve semantic issues, and the word processing stage, especially to the sentence processing stage, the understanding of words, phrases and sentences need to generate a variety of semantic knowledge, including the semantic classification of words, Semantic collocation features and semantic roles and other aspects of the detailed description. The results of Mongolian contemporary semantics research largely bias the abstract explanation of semantic concepts and neglect the specific expressions of the outer semantic. So it can not meet the actual needs of information processing. The majority of the Mongolian semantic knowledge is explained in the syntactic structure. However, the syntactic analysis of Mongolian information processing at the present stage is an urgent issue to be solved. Taking the Mongolian high-frequency words and the real corpus as the main material, this paper makes exploratory researches on semantic classification, semantic field division, sememe analysis, valence and grammar of grading, and obtains some cognitions, and masters some practical and effective Semantic analysis method.