2026-05-09T11:14:53.906Z http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai

ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai:3:11212 2025-03-01T10:04:20Z 3

Job Satisfaction Analysis: Optimizing eXtreme Gradient Boosting with Natural Language Processing Through Modified Metaheuristic, Chapter in SIST Smart Innovation, Systems and Technologies: BIDA 2024: Business Intelligence and Data Analytics, Springer, volume 413 2025 http://ezaposleni.singidunum.ac.rs/rest/sciNaucniRezultati/oai/record/3/11212 https://link.springer.com/chapter/10.1007/978-981-97-7717-4_30 A. Bozovic L. Jovanovic A. Nastasic M. Antonijevic N. Bacanin M. Zivkovic This research explores the applicability of artificial intelligence to analyze and interpret employee feedback in the context of job satisfaction. Recognizing the pivotal role that employee feedback plays as an indicator of work culture-integral for effective talent acquisition and retention - the study employs frequency-inverse document frequency encoding to decipher key terms embedded in provided feedback. The subsequent step involves subjecting this information to classification through the XGBoost algorithm, a powerful tool for predictive modeling. Acknowledging that the performance of the XGBoost algorithm is contingent upon well-considered hyperparameter selections, the research employs a diverse set of optimizers to enhance its efficacy. In response to identified shortcomings, a modified version of an optimizer is introduced, aiming to refine the algorithm’s performance. The research acknowledges the inherent challenges of the task, resulting in demonstrated results that, while modest, shed light on the complexities involved in interpreting employee feedback. Despite the inherent challenges, the introduced algorithm exhibits notable performance superiority when compared to other optimizers. This finding underscores its potential significance in refining the interpretation of employee feedback within workplace dynamics. The study contributes to the evolving landscape of AI applications in the realm of employee satisfaction assessment, offering insights and opportunities for further advancements in this critical domain. bookPart Springer, Singapore 419 434 10.1007/978-981-97-7717-4_30 SIST Smart Innovation, Systems and Technologies: BIDA 2024: Proceedings of Business Intelligence and Data Analytics, volume 413