A Topographical Sentimental Analysis of COVID-19 Tweets

Abstract: The spread of Covid-19 has resulted in worldwide health concerns. Social media is increasingly used to share news and opinions about it. Therefore, it is the need of the hour to implement different measures to safeguard the countries by demystifying the pertinent facts and information. This paper focused on sentiment analysis on Twitter data related to Covid-19 using RNN. Originally, the input data collected from the dataset are pre-processed and applied to the sentiment lexicon dictionary for estimating sentiment values. The output data with the sentiment values are clustered using the Probability-based K- medoid algorithm to find whether the tweet caries any sentiment or not. Then from clustered results, the important is extracted and tagged using Part of Speech (POS). Then the Optima POS words are selected by using the Uniform Distribution based Cat Swarm Optimization (UD-CSO) algorithm and are subjected to Xavior initialization and Linear Logistic based Recurrent Neural Network (XL2-RNN) for classification. The experimental results indicated the effectiveness of the proposed sentiment analysis model.

Keywords: Probability-based K- medoid algorithm, Uniform Distribution based Cat Swarm Optimization (UD-CSO), Part of Speech (POS), Xavior initialization and Linear Logistic based Recurrent Neural Network (XL2-RNN), Covid-19 tweets, Sentiment Analysis