An Intelligent Multimodal Sentiment Analysis and Emotion Detection Framwork for Multimedia Reviews

Abstract

ii newlineABSTRACT newlineThe data revolution is one of the most important features of this century. In the last two newlinedecades, due to the availability of low-cost Internet facilities, advancements in Word- newlineWide-Web, and the growth in social media sites, multimodal data has grown at an newlineexponential rate. Multimodal data analysis can be used in many fields like improving the newlinequality of life of a customer, improving the business by monitoring the perception of newlineconsumers towards products or services or brands, human-computer interaction, TV news newlinesummarization, online interview, student feedback analysis, etc. The extraction of newlinecontextual knowledge from the utterances of a video, efficient multimodal fusion, feature newlinealignment between multiple modalities, handling conversational videos, and optimal newlinefeature selection are important issues in multimodal sentiment analysis and emotion newlinedetection.