Volume 18, No. 6, 2021

An Approach Of Multimodal Sentiment Analysis Using Machine Learning

Omprakash Dewangan , Dr. Megha Mishra


Our reliance on social media has grown as a result of the widespread use of the Internet. It's no secret that Twitter is one of the most widely used social media platforms today. People express their views on a wide range of topics via the medium of Twitter, including politics, sports, the economy, and so on. The sheer volume of data generated by millions of tweets every day grabbed the attention of data scientists, who began analysing it for the sentiment. Social media postings are categorized as either good, negative, or neutral using the sentiment analysis method. As a result, the research intends to examine how different forms of text representation affect sentiment analysis performance. The studies in this research employed data from two different datasets. First, Kotzias tagged movie reviews from the Internet Movie Database (IMDB), MOSI, MOUD, and then he gathered tweets on health-related themes in English using the Twitter API. Researchers employed Naive Bayes, Support Vector Machines, and Artificial Neural Network approaches to develop classification models and Python was used to classify the attitudes as positive, negative, or neutral. TF-IDF and Word2Vec (W2V) modelling approaches were used to extract features from the dataset. In the conclusion, the categorization algorithms' success rates were compared. The results of the experiments showed that Artificial Neural Networks performed the best in terms of accuracy in both datasets.

Pages: 8491-8503

Keywords: Artificial Neural Network, Machine Learning, Naive Bayes, Sentiment analysis, Social Support Vector Machines.

Full Text