《计算机应用研究》|Application Research of Computers

多特征融合的图文微博情感分析

Multimedia sentiment analysis on microblog based on multi-feature fusion

免费全文下载 (已被下载 次)  
获取PDF全文
作者 凌海彬,缪裕青,张万桢,周明,武继刚
机构 1.桂林电子科技大学 a.计算机与信息安全学院;b.图像图形智能处理重点实验室,广西 桂林 541004;2.桂林航天工业学院 实践教学部,广西 桂林 541004;3.桂林海威科技股份有限公司,广西 桂林 541004;4.广东工业大学 计算机科学与技术学院,广州 510006
统计 摘要被查看 次,已被下载
文章编号 1001-3695(2020)07-004-1935-05
DOI 10.19734/j.issn.1001-3695.2018.12.0929
摘要 现有的微博情感分析方法已经注意到了微博文本与图片之间的互补作用,但较少注意用户情感表达的差异和微博内容中除文字之外的特征,为此提出一种多特征融合的图文微博情感分析方法。首先构建文本情感分类模型,将对情感具有很好指示作用的内容特征和用户特征与微博句子进行融合, 然后构造了基于参数迁移和微调的图片情感分类模型。最后设计特征层和决策层融合的方法,将文本和图片情感分类模型进行融合。实验结果表明,内容特征和用户特征有效增强了模型捕捉情感语义的能力,并在多项性能指标上都取得了很好的效果, 构建的图文情感分类模型和融合方法可获得更好的性能。
关键词 情感分析; 微博; 多特征融合; 神经网络; 图文融合
基金项目 国家自然科学基金资助项目(61763007,61866007)
广西自然科学基金联合资助项目(2018GXNSFAA138082)
广西高校图像图形智能处理重点实验室资助项目(GIIP201706)
桂林电子科技大学研究生教育创新计划基金资助项目(2017YJCX50)
本文URL http://www.arocmag.com/article/01-2020-07-004.html
英文标题 Multimedia sentiment analysis on microblog based on multi-feature fusion
作者英文名 Ling Haibin, Miao Yuqing, Zhang Wanzhen, Zhou Ming, Wu Jigang
机构英文名 1.a.School of Computer Science & Information Security,b.Key Laboratory of Image & Graphics Intelligent Processing,Guilin University of Electronic Technology,Guilin Guangxi 541004,China;2.Practice Teaching Department,Guilin University of Aerospace Technology,Guilin Guangxi 541004,China;3.Guilin Hivision Technology Company,Guilin Guangxi 541004,China;4.School of Computer Science & Technology,Guangdong University of Technology,Guangzhou 510006,China
英文摘要 Previous studies exploit the complementary effect between text and images in microblog sentiment analysis, but less focuse on some factors such as user personality and content characteristics. To solve this problem, this paper proposed an multimedia sentiment analysis on microblog based on multi-feature fusion method. Firstly it constructed a text sentiment classification model considering the user-based features and content-based features which had a good indication of emotions. Then it constructed an image model based on parameter transfer and fine-tuned. Finally, it designed fusion method of the early fusion and the late fusion. The experimental results show that user-based features and content-based features can enhance model ability of capturing emotion semantic, and the proposed model and fusion methods get higher accuracy on the real-world test dataset.
英文关键词 sentiment analysis; microblog; multi-feature fusion; neutral network; text and image
参考文献 查看稿件参考文献
 
收稿日期 2018/12/14
修回日期 2019/3/7
页码 1935-1939,1951
中图分类号 TP391
文献标志码 A