计算机应用 ›› 2021, Vol. 41 ›› Issue (5): 1256-1261.DOI: 10.11772/j.issn.1001-9081.2020081242

所属专题: 人工智能

• 人工智能 • 上一篇    下一篇

基于生成对抗网络的事件描述生成

孙鹤立1, 孙玉柱1,2, 张晓云1   

  1. 1. 西安交通大学 计算机科学与技术学院, 西安 710049;
    2. 西安交通大学 外国语学院, 西安 710049
  • 收稿日期:2020-08-18 修回日期:2020-10-14 出版日期:2021-05-10 发布日期:2020-12-09
  • 通讯作者: 孙玉柱
  • 作者简介:孙鹤立(1983-),女,辽宁铁岭人,副教授,博士,CCF会员,主要研究方向:数据挖掘、城市计算;孙玉柱(1982-),男,河北深州人,工程师,博士研究生,主要研究方向:数据挖掘、机器学习;张晓云(1996-),男,江苏苏州人,硕士研究生,主要研究方向:数据挖掘、机器学习。
  • 基金资助:
    国家自然科学基金资助项目(61672417)。

Event description generation based on generative adversarial network

SUN Heli1, SUN Yuzhu1,2, ZHANG Xiaoyun1   

  1. 1. School of Computer Science and Technology, Xi'an Jiaotong University, Xi'an Shaanxi 710049, China;
    2. School of Foreign Studies, Xi'an Jiaotong University, Xi'an Shaanxi 710049, China
  • Received:2020-08-18 Revised:2020-10-14 Online:2021-05-10 Published:2020-12-09
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61672417).

摘要: 在基于事件的社会网络(EBSN)中,自动生成社交事件(Social Event)的事件描述(Event Description)供组织者参考,从而有效避免描述贫乏、描述过度、精准度低的问题,易于形成丰富、准确、高吸引力的事件描述。为了自动生成与真实事件描述足够相似的文本,提出了一种生成对抗网络(GAN)模型GAN_PG来生成事件描述。GAN_PG模型中的生成模型(Generator)采用变分自编码器(VAE),判别模型(Discriminator)采用带门控循环单元(GRU)的神经网络。模型训练时借鉴了强化学习中的策略梯度(PG)下降,并通过设计合理的奖励函数来训练生成器生成事件描述。实验结果表明,设计的模型生成事件描述的BLEU-4值达到了0.67,证明了提出的事件描述生成模型GAN_PG可以无监督地产生与自然语言足够相似的事件描述。

关键词: 基于事件的社会网络, 事件描述, 文本生成, 生成对抗网络, 变分自编码器

Abstract: In Event-Based Social Networks (EBSNs), generating the event description of social events automatically is helpful for the organizer, so as to avoid the problems of poor description, descripting too much and low accuracy, and be easy to form rich, accurate and attractive event description. In order to automatically generate text that is sufficiently similar to true event description, a Generative Adversarial Network (GAN) model named GAN_PG was proposed to generate event description. In the GAN_PG model, the Variational Auto-Encoder (VAE) was used as the generator, and the neural network with the Gated Recurrent Unit (GRU) was used as the discriminator. In the model training, the Policy Gradient (PG) decline in reinforcement learning was used as reference, and a reasonable reward function was designed to train the generator to generate event description. Experimental results showed that the BLEU-4 value of the event description generated by GAN_PG reached 0.67, which proved that the event description generation model GAN_PG can generate event descriptions sufficiently similar to natural language in an unsupervised way.

Key words: Event-Based Social Network (EBSN), event description, text generation, Generative Adversarial Network (GAN), Variational Auto-Encoder (VAE)

中图分类号: