Analysis of the Effect of Compensation on Twitter Based on Job Satisfaction on Sustainable Development of Employees Using Data Mining Methods
Sandeep Kumar Gupta, P Reznik Nadia, Esra Sipahi Döngül, Sayonara de Fátima Teston, Abrham Fantaw · Journal of Talent Development and Excellence · 2020
It is very important that the managers responsible for the performance of their employees know the reasons that affect their job satisfaction in order to carry the organizations to the future and continuity. From this point, the involvement of employees in the development of the organization, the achievement or dissatisfaction of loyalties depends on the people in the Organization. It has shown on various events that one of the ways to deal with get the best execution out of agents is to make them pleasant and perky. Workers everything being equivalent and levels of pay continue getting dynamically hopeless at work - an extended stretch example that should concern supervisors. How, by then do directors make satisfied agents? Since pay has been a fundamental thought of pushing delegates in an Organization. The present investigation broke down the issue of pay and compensation based satisfaction. Data was accumulated and separated similar to drawing in experiences, and the analysis of the data was done through the SPSS package program with a 95% confidence interval. Chi-square test was used to investigate with frequencies and explanatory information. The other purpose of this study is to examine the tweets of people who use social media, based on their own declarations, who have a high job satisfaction due to their salary, and those who work with less salary, and as a result of the analysis on tweets sent by these people, the employees who love their job are distinctive. to make clear differences. In this study, data mining, machine learning and data science methods are applied to data captured on Twitter. A total of 142,656 tweets were made and these tweets were worked on. Among the applied algorithms, the highest success rate belongs to the Grandient Boosted Tree and it can accurately classify on a nearly balanced and bipolar dataset with a success over 99%. As a result of the applied machine learning algorithms, it was observed that there was a significant difference between the sentence structures, the words used and the posts made by those who were satisfied and dissatisfied with the salary they received. Thanks to machine learning, people who have high job satisfaction and who do not have high job satisfaction can be learned by analyzing their twitter accounts. Other machine learning algorithms can also be attempted when sufficient hardware capacity is available. Techniques have used to test the importance of association among pay and compensation based satisfaction. The result demonstrated that there is no significant association among compensation and delegate occupation satisfaction among the respondents. In any case portions of pay, for instance, livelihood improvement and business dependability were significant contributing factors to delegate occupation satisfaction. The assessment recommended that better conveyor headway openings should be given to the agents to assemble work satisfaction.