处理复杂信息分类的自然语义模型仿真分析

Analysis of Natural Semantic Model Simulation for Processing Complex Information Classification

在对复杂信息分类问题进行处理时,由于关键词的多义性,导致传统的复杂信息分类方法在信息分类时存在一定的混淆,无法准确完成分类。提出一种处理复杂信息分类的自然语义模型,依据复杂信息决策表和复杂信息决策规划集,对复杂信息分类进行预处理,通过塑造关键词分类词典、同义词词典、经切词、网页文档扫描统计完成复杂信息分类文档特征关键词的抽取,经过统计分析获取所有关键词在不同专题中的隶属度,将其组成模糊关键词集合,利用学习机制获取复杂信息分类特征指标和专题类别的关联度,采用梯度下降法对复杂信息分类参数进行训练。仿真实验结果表明,所提方法具有很高的准确性。

In the classification problem of complex information processing, because of the ambiguity keywords, lead to thetraditional complex information classification method has certain confusion when information classification, can＇t do classi-fication accurately. Put forward a kind of natural semantic model of processing complex information classification based oncomplex information program sets, decision tables, and complex information preprocessing, classification of complex infor-mation through shape classification dictionary, thesaurus, the word, web document scan statistics accomplish complex infor-mation classification of the characteristics of document keywords extraction, through statistical analysis to obtain all themembership degree of keywords on different topics, the fuzzy key word set, using the learning mechanism for complex infor-mation classification characteristic index and correlation of the thematic categories, using the gradient descent method forcomplex information classification parameters for training. The simulation results show that the proposed method has highaccuracy.