MULTIMODAL SENTIMENT ANALYSIS

MULTIMODAL SENTIMENT ANALYSIS

Akshada Kothavale¹, Khushi Sahoo², Kedar Awale³, Dnyanesh Kunkulol⁴ , Priyanka Khalate⁵

^1,2,3,4Student, Dept. of Computer Engineering, S.K.N.C.O.E. Vadgaon, Pune, Maharashtra, India

(⁵Professor, Dept. of Computer Engineering, S.K.N.C.O.E. Vadgaon, Pune, Maharashtra, India)

Abstract - In this article, we provide a novel story of multimodal sentiment analysis that includes the collection of sentiments from Online videos using a model that incorporates audio, visual, and textual modalities as knowledge resources. We employed feature- of new neural network architectures, including as autoencoder networks, convolutional neural networks (CNNs), deep belief networks (DBNs), and memory-enhanced neural network models such as long short-term memory (LSTM) models, have lately acquired prominence.

Keywords — Sentiment Analysis, Multimodal Sentiment Analysis, Multimodal Fusion, Human Computer Interaction