论文部分内容阅读
文章提出了一个利用大规模真实语料自动提取汉语词汇联想意义的方法,利用“像Y那样A”和“A得像Y那样”作为提取框架作连续推演,将生成一个体词和谓词构成的复杂关系系统。提取的联想意义构成了一个封闭而又无限的神经网络,网络每一个节点的“意义”实际上是由该节点所连接的弧线所定义,节点词语的意义就是各弧线另一端点意象的复合,这种意义是真实约,而传统义素分析法所流行的二分树形的义素组合则是虚拟的,而且具有很大的随意性。文章构建了一个汉语词汇联想意义客观的、量化的、系统的多维模型。
In this paper, a method of automatically extracting associations of Chinese words using large-scale real corpora is proposed. Using “A like Y” and “A like Y” as a continuous deduction framework, a complex sentence formed by body and predicate Relationship system. The extracted meaning of association constitutes a closed and infinite neural network. The meaning of each node in the network is actually defined by the arc connected to the node. The meaning of the node words is the image of the other end point of each arc Complex, this sense is true about, and the traditional bizarre analysis of the popular dichotomy of the semantic combinations is virtual, and has a lot of randomness. The article constructs an objective, quantitative and systematic multidimensional model of association of Chinese words.