Prediction model for assessing the Sri Lankan employees' perception on remote working in post-covid-19 pandemic

Abstract

It is important to protect all employees from contacting COVID-19 since they want to continue their careers. Therefore, the remote working process has been sped up for continuing the work without any interruptions during this Covid- 19 pandemic. Many organizations have been persuaded by the pandemic that remote work has benefits for a successful organization. The objective of the study is to identify the employees’ ideas and assess their sense for continuing the remote working concept in the post-COVID-19 pandemic. The study used a random sample of 325 workers in private and public sectors who did remote work during the pandemic and collected data via questionnaires through Google form. Information gain ranking was applied to the ranking of attributes and removed unnecessary data. After the pre-processing, the prediction models were generated using the WEKA tool, and 66% of the percentage split used with six different classification algorithms such as Support Vector Machine, Naive Bayes, Logistic Regression, Decision Tree, Multi-Layer Perceptron, and Random Forest and ensemble learning algorithm that combined above six algorithms using an average of probabilities through vote algorithm. Based on the evaluation results of accuracy, recall, precision, fmeasure, and error values, the ensemble learning algorithm outperforms the other six algorithms by 90%. According to the data set, the majority of employees prefer the remote working concept. Using this prediction model, we can also assess the employee's sense and compatibility for the remote working concept. In the future, we plan to expand the data set and conduct more evaluations using deep learning.