Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Visual and text sentiment analysis through hierarchical deep learning networks
Chaudhuri A., Springer International Publishing, New York, NY, 2019. 120 pp. Type: Book
Date Reviewed: Jan 25 2021

This book is on the extraction of sentiments from text/image data using machine learning. It describes research related to developing a deep learning technique for the extraction. The technique uses hierarchical gated feedback recurrent neural networks (HGFRNNs).

Chapter 1 provides motivation for the research work: a sentiment extraction method, an evaluation of the deep learning algorithm involved, and an evaluation of multimodal data analysis.

Chapters 2 and 3 present a survey of literature on sentiment extraction methods and a survey of the available technologies for such applications. Chapter 4 is on the datasets used in the experiments. The datasets were taken from social media, such as Twitter, Instagram, Viber, and Snapchat. Each dataset contain thousands of images and millions of text messages. Chapter 5 briefly discusses sentiment extraction from images.

Chapter 6’s five sections cover neural networks. Section 6.1, “Deep Learning Networks,” gives a mathematical introduction to recurrent neural networks (RNNs) and long short-term memory (LSTM) networks. Section 3 describes the time dynamics of gated feedback RNN (GFRNNs), which is an extension of RNN. Sections 4 and 5 discuss HGFRNNs in detail, including mathematical abstraction, the layers with respect to forward/backward passes, and multimodal aspects. The chapter presents the involved problems, and solves them.

Chapter 7 uses a baseline method, the cross-media bag-of-words model (CBM), to describe the experimental results. For each dataset, the accuracy of the HGFRNN results is compared with the accuracy of the CBM results. The sentiment analysis results are compared for text-only data, image-only data, and fusion data. The analysis also includes error analysis for each dataset. The last chapter concludes: HGFRNNs give better results. The appendix provides sample images from various datasets.

The book represents specific research and thus cannot be used as a textbook. Readers interested in sentiment analysis research will find it useful. The research is a good contribution to our understanding of HGFRNNs and the development of a technique for sentiment analysis.

Reviewer:  Maulik A. Dave Review #: CR147168 (2106-0141)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Neural Nets (I.5.1 ... )
 
 
Hierarchical (I.4.10 ... )
 
 
General (H.0 )
 
Would you recommend this review?
yes
no
Other reviews under "Neural Nets": Date
Synergetic computers and cognition
Haken H. (ed), Springer-Verlag New York, Inc., New York, NY, 1991. Type: Book (9780387530307)
Oct 1 1992
Code recognition and set selection with neural networks
Jeffries C., Birkhäuser Boston Inc., Cambridge, MA, 1991. Type: Book (9780817635855)
Jun 1 1993
Fast learning and invariant object recognition
Souček B. (ed), Wiley-Interscience, New York, NY, 1992. Type: Book (9780471574309)
Nov 1 1992
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy