基于改进rcnn模型的多标签短文本自适应分类  

Multi-Label Short Text Adaptive Classification Based on Improved Rcnn Model

作　　者：王东[1] 夏梓渊 WANG Dong;XIA Zi-yuan(School of Computer Science and Engineering,Chongqing University of Technology,Chongqing 400054 China)

机构地区：[1]重庆理工大学计算机科学与工程学院,重庆400054

出　　处：《计算机仿真》2021年第5期388-392,共5页Computer Simulation

摘　　要：为了提升传统多标签短文本自适应分类方法的分类准确率,提出基于改进rcnn模型的多标签短文本自适应分类方法。首先在多标签短文本数据集中提取多标签短文本的不同特征,将其作为传统机器学习模型以及深度模型的输入;然后结合Stacking技术对rcnn模型模型进行改进,通过改进的rcnn模型对多个基分类器的分类结果进行融合处理,获取多标签短文本自适应分类的最终结果。最后进行仿真测试,结果表明,所提方法能够快速、准确完成多标签短文本自适应分类。In order to improve the classification accuracy of the traditional multi-label short text adaptive classification method, a multi-label short text adaptive classification method based on the improved rcnn model was proposed. First, the different features of the multi-label short text were extracted from the multi-label short text data set, and they were used as the input of traditional machine learning models and deep models. Then, the rcnn model was improved with Stacking technology. The classification results of the classifier were fused to obtain the final result of adaptive multi-label short text classification. Finally, simulation experiments were performed. The results of the simulation experiments show that the proposed method can quickly and accurately complete multi-label short text adaptive classification.

关 键 词：多标签短文本 自适应分类 仿真 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]