Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A novel system for video news' sentiment analysis

A novel system for video news' sentiment analysis Purpose – Despite the actual prevalence of diverse types of multimedia information, research on video news is still in an early stage. Improving the accessibility of video news seems worth investigating, therefore, the purpose of this paper is to present a new combination mode of video news text clustering and selection. This method is useful for sorting out and classifying various types of news videos and media texts based on sentiment analysis. Design/methodology/approach – A novel system is proposed, whereby video news are identified and categorized into good or bad ones via the authors' suggested Hidden Markov Model (HMM) and Support Vector Machine (SVM) hybrid learning method. Actually, an exploratory video news sentiment analysis case study, conducted on various news databases, has proven that the feature‐selection‐combining method, encompassing the Information Gain (IG), Mutual Information (MI) and CHI‐statistic (CHI), performs the best classification, which testifies and highlights the designed framework's value. Findings – In fact, the system turns out to be applicable to several areas, especially video news, where annotation and personal perspectives affect the accuracy aspect. Research limitations/implications – The present work shows the way for further research pertaining to the personal attitudes and the application of different linguistic techniques during the classification. Originality/value – The achieved results are so promising, encouraging and satisfactory, that they highlight the originality and efficiency of the authors' approach as an effective tool enabling to secure an easy access to video news and multi‐media texts. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Systems and Information Technology Emerald Publishing

Loading next page...
 
/lp/emerald-publishing/a-novel-system-for-video-news-sentiment-analysis-Z9h0zdCeS3

References (37)

Publisher
Emerald Publishing
Copyright
Copyright © 2013 Emerald Group Publishing Limited. All rights reserved.
ISSN
1328-7265
DOI
10.1108/13287261311322576
Publisher site
See Article on Publisher Site

Abstract

Purpose – Despite the actual prevalence of diverse types of multimedia information, research on video news is still in an early stage. Improving the accessibility of video news seems worth investigating, therefore, the purpose of this paper is to present a new combination mode of video news text clustering and selection. This method is useful for sorting out and classifying various types of news videos and media texts based on sentiment analysis. Design/methodology/approach – A novel system is proposed, whereby video news are identified and categorized into good or bad ones via the authors' suggested Hidden Markov Model (HMM) and Support Vector Machine (SVM) hybrid learning method. Actually, an exploratory video news sentiment analysis case study, conducted on various news databases, has proven that the feature‐selection‐combining method, encompassing the Information Gain (IG), Mutual Information (MI) and CHI‐statistic (CHI), performs the best classification, which testifies and highlights the designed framework's value. Findings – In fact, the system turns out to be applicable to several areas, especially video news, where annotation and personal perspectives affect the accuracy aspect. Research limitations/implications – The present work shows the way for further research pertaining to the personal attitudes and the application of different linguistic techniques during the classification. Originality/value – The achieved results are so promising, encouraging and satisfactory, that they highlight the originality and efficiency of the authors' approach as an effective tool enabling to secure an easy access to video news and multi‐media texts.

Journal

Journal of Systems and Information TechnologyEmerald Publishing

Published: Mar 15, 2013

Keywords: Video; Multimedia; Information media; Mass media; Video news; Text detection; Annotation; Combination of feature‐selection information; Sentiment analysis

There are no references for this article.