基于SVM金融分析报告的行业分类研究

针对传统金融分析报告分类效率低的问题，提出基于支持向量机的中文文本分类技术来对金融分析报告进行分类，该分类技术采用中科院提供的中文分词系统以及使用两种特征选择算法相结合进行分词和特征选择，并且提出针对TF／IDF权重计算的改进方法。该分类技术选择支持向量机作为分类算法，通过开源的支持向量机对样本进行训练和测试。实验结果表明，采用中文文本分类技术对金融分析报告按照行业进行分类能够满足金融机构的使用需求。

Research on the Classification of Financial Analysis Report Based on SVM

In view of the problem of low efficiency on classification in traditional financial analysis report, presents text classification based on support vector machine （SVM） to classify financial analysis report. It makes word segmentation experient by using the Chinese word segmentation system provided by the Chinese Academy of Sciences and used the combining method of two feature selection algorithms to test. Then it comes up a TF/IDF weighting computation method. Choses support vector machine as a classification algorithm, finally it takes sample training and testing through the open source of support vector machine. In the end, the experiental result shows that the Chinese text classification technology can meet the demand of the financial institutions by classifying financial analysis report according to industry classification.