非结构化网络中有价值信息数据挖掘研究

摘要

A data mining method of valuable information in the unstructured network is proposed based on the association rules.Firstly,the feature extraction method is used to classify and recognize the text feature preliminarily,and the feature of different text types is extracted.Then,the association rules method is used to calculate the association degree among each types of feature,and the more obvious word in the feature of different text types is used as the clustering object.Moreover,the document classification scheme with the maximum posterior probability on the given sample data is searched out,and the information of every document is used as the random symbol sequence formed by its category.Finally,the maximum likelihood estimation is used to work out the frequency distribution of each symbol,and the calculation result is used to complete the information data mining.The simulation results show that the method has good expandability.It can complete the data mining of valuable information in the unstructured network effectively.%对非结构化网络中有价值信息数据挖掘的研究,可更好的提升非结构化海量网服务质量.采用当前方法进行数据挖掘时,需要对信息数据先进行降噪,但降噪过程较为复杂,挖掘方法局限性大.提出基于关联规则的非结构化网络中有价值信息数据挖掘方法.利用特征提取方法对文本特征进行初步的分类与识别,提取不同文本类型的特征,利用关联规则方法计算各个类型特征间的关联度,将不同文本类型特征中较为明显的词作为聚类的对象,搜索出在给定样本数据上具有最大后验概率的文档划分方案,将每个文档信息作为其类别形成的随机符号序列,采用最大似然估计计算出各个符号的频率分布,利用计算的结果完成对信息数据挖掘.仿真结果表明,所提方法可扩展性较强,可以有效地完成对非结构化网络中有价值信息数据挖掘.