文本情报信息筛选与聚类的一种处理方法

目前大数据时代情况下的信息有文本、图像、语音和视频等多种形式,而且信息的容量非常大,怎样高效、正确地筛选、分类和处理、利用这些信息,为决策者提供指挥与控制的科学依据显得尤为重要。据此对文本情报信息提出了一种文本聚类的特征选择以及特征变换的方法,利用单词在文本中的出现次数的概率来选择参与聚类的单词,并且对单词出现概率模型定义了特征变换函数,提高了文本信息的筛选、分类和处理的精度,能快速、准确地提取所需要的情报信息提供给指挥与控制的决策者参考、使用。

A Text Clustering Method Using Word Appearance Probability

Currently,in the big data era,information includes text,image,voice and video etc, and the volume of information is extremely large. So how to filter,classify,processing and use these information efficiently,and offering support for command and control becomes very import. As this, the proposed method which is a feature selection and feature transform method, uses the"word appearance probability"to select which term will be used as a feature while clustering and then convert word appearance frequency to a value calculated by a proposed function which defined with"word appearance probability". The precision of filtering classify and processing for text information is improved,which can offering the information required by decision-maker fast and precise