基于LDA的主题分类系统研究

Study on the topic classification system based on LDA

当前人类处于信息爆炸的时代,对于海量的文本数据,可以利用人工智能的工具来提高数据分析处理的效率,来挖掘海量数据的宝藏。文章主要对文本的主题分类算法进行研究,通过改进分类方法并提出可视化方案,使主题分类具有更好的应用价值。首先通过利用LDA主题分类算法进行处理,并提出了一些改进方法使分类效果更优,并最终生成可视化的主题分类结果,进而用于推荐系统、数据挖掘、数据分析等领域。

At present, human beings are in the era of information explosion. For massive text data, artificial intelligence tools can be used to improve the efficiency of data analysis and processing, and to excavate the treasure of massive data. This paper mainly focuses on the research of text topic classification algorithm. By improving the classification method and putting forward the visualization scheme, the topic classification has better application value. First, LDA theme classification algorithm is applied to deal with it, and some improvement methods are put forward to make the classification effect better, and finally generate visual subject classification results, which is further applied to the fields of recommender system, data mining, data analysis and so on.