电信科学 ›› 2018, Vol. 34 ›› Issue (12): 117-124.doi: 10.11959/j.issn.1000-0801.2018301

• 专栏:人工智能技术与应用 • 上一篇    下一篇

基于深度学习的文本分类系统关键技术研究与模型验证

汪少敏,杨迪,任华   

  1. 中国电信股份有限公司上海研究院,上海 200122
  • 修回日期:2018-12-10 出版日期:2018-12-01 发布日期:2019-01-02
  • 作者简介:汪少敏(1983-),女,中国电信股份有限公司上海研究院高级工程师,主要研究方向为人工智能技术、自然语言处理、大数据和数据挖掘分析。|杨迪(1982-),男,中国电信股份有限公司上海研究院高级工程师,主要研究方向为人工智能技术、智能交互。|任华(1977-),女,中国电信股份有限公司上海研究院高级工程师,主要研究方向人工智能技术、智能客服。

Key technology research and model validation of text classification system based on deep learning

Shaomin WANG,Di YANG,Hua REN   

  1. Shanghai Research Institute of China Telecom Co.,Ltd.,Shanghai 200122,China
  • Revised:2018-12-10 Online:2018-12-01 Published:2019-01-02

摘要:

大数据时代,文本分类是文本数据挖掘和文本价值探索领域的重要工作。传统的文本分类系统存在特征提取能力弱、分类准确率不高的问题。相对于传统的文本分类技术,深度学习技术具有准确率高、特征提取有效等诸多优势,有必要将深度学习技术引入文本分类系统,以解决传统文本分类系统存在的问题。在分析传统文本分类系统的基础上,提出了基于深度学习的文本分类系统的体系架构和关键技术,同时对传统分类模型、TextCNN、CNN+LSTM多种分类模型进行了验证比对。

关键词: 深度学习, 文本分类, 分类模型, 神经网络

Abstract:

Text classification is very important to text data mining and value exploration.The traditional text classification system has problems of weak feature extraction ability and low classification accuracy.Compared with the traditional text classification technology,deep learning technology has many advantages such as high accuracy and effective feature extraction.Therefore,it is necessary to apply deep learning technology to the text classification system to solve the problems of the traditional text classification system.The traditional text classification system was analyzed,and the architecture and key technologies of text classification system based on deep learning were proposed.Finally,several classification models were verified and compared,including the traditional classification model,TextCNN and CNN+LSTM.

Key words: deep learning, text classification, classification model, neural network

中图分类号: 

No Suggested Reading articles found!