摘要
通过构建有限领域概念网模型,形成一种新颖的自然语言文本处理方法。该方法首先构建领域概念网,再对目标文本进行常规处理后提取其语句框架,最后计算文本相似度。据此开发的试卷主观题评测系统表明,与目前方法相比,该系统以语义为核心,使得命题所指的概念网与考生答案的语义之间有联系,从而提高评分准确率。
By constructing a finite domain network of concepts model, a novel form of natural language text processing methods.Firstly, the method build domain network of concepts, and then regularly handle with the target text frame, finally calculate the text similarity. According to the subjective question papers developed evaluation systems says that, compared with current methods, this system semantics as the core, to keep within the meaning of the network of concepts propositions and candidates answer semantic links, improving the score accuracy.
出处
《电脑知识与技术（过刊）》
2014年第10X期6827-6830,共4页
Computer Knowledge and Technology
关键词
概念网
语句框架
词网
句子相似度
主观题评分
the network of concepts
sentence framework
wordnet
sentence similarity
assessment of subjective questions
