Sentimental Analysis (SA) of Employee Job Satisfaction from Twitter Message Using Flair Pytorch (FP) Method

Abstract

Organizations in the contemporary period face a number of problems as a result of the changing nature of the environment. One of a company's numerous problems is to please its workers in order to manage with an ever-altering and dynamic environment, attain success and stay competitive. The firm must meet the demands of its employees by offering appropriate working circumstances for enhancing efficacy, proficiency, job dedication, and throughput. Twitter is an online social networking site where users may share their thoughts on a wide range of topics, debate current events, criticize and express a wide range of feelings. As a result, Twitter is one of the greatest sources of information for emotion analysis, sentiment analysis, and opinion mining. Owing to a huge volume of opinionated material now created by internet users, Sentiment Analysis (SA) has emerged highly popular in both industry and research. Thus, this paper examines the problem by examining sentiment text as well as emotion symbols such as emoji. Therefore, utilizing the Flair Pytorch (FP) technique, an embedding type Natural Language Programming (NLP) system, and unique strategy for Twitter SA with a focus on emoji is presented. It overtakes state-of-the-art algorithms when it comes to pulling out sentiment aware implanting of emoji and text. Furthermore, 3520 tweets from an organization are accumulated as a dataset, with each tweet containing an emoji. As a result, the recommended FP technique has utilized the “en-sentiment” model for text classification and tokenization to determine the divergence of a sentence established on sentiment words, such as negative or positive, in the sentimental status of tweet, which could be assessed using the respective method’s confidence score.