CAJ | 학술논문

动词+名词现象在中文句法分析中具有非常重要的地位,它直接影响着句子的结构分析,隐马尔科夫支持向量机是兼顾词语前后关系的一种结构化学习模型,该模型在英文句法分析中的有效性已经被证明,将该模型用于中文动词+名词关系分析识别任务,实验结果表明,该模型能获得很好的关系识别效果.
동사+명사현상재중문구법분석중구유비상중요적지위,타직접영향착구자적결구분석,은마이과부지지향량궤시겸고사어전후관계적일충결구화학습모형,해모형재영문구법분석중적유효성이경피증명,장해모형용우중문동사+명사관계분석식별임무,실험결과표명,해모형능획득흔호적관계식별효과.
Verb+noun relation is an important structure in Chinese parsing,it has effect on analysing sentence structure.Hidden Markov Support Vector Machines is a novel structural learning model that can take account of the neighboring relations before or after the current words. Its efficiency has been proved in English parsing. In this paper, we apply HMSVM to solve Chinese verb+noun relation problem. The results of experiment show a better relation recognition effect.