【摘要】 格机是一种新颖的有监督学习方法。论文扩展了格机的等标注概念,提出了交集标注的概念,以解决决策系统中的同条件多决策值问题,并将其用于非结构化的文本分类领域。作者用自动抽取的代表文档特征的关键词表示文档,用交集标注缩减数据,对新文本作分类预测。给出了求交集标注的算法。实验表明扩展后的格机是很好的文本分类学习方法。更多还原

【Abstract】 Lattice machine is a novel approach to supervised learning.In this paper,author extend the concept of″equil-abeled″in lattice machine,present ″intersection labeled″to solve the problem of multi-decision attribute values for same condition in a decision system,and investigate the application of the intersection labeled in the area of unstruc-tured text classification.The author represent a text with更多还原