TY - GEN
T1 - Employee Satisfaction in Online Reviews
AU - Koncar, Philipp
AU - Helic, Denis
PY - 2020/1/1
Y1 - 2020/1/1
N2 - Employee satisfaction impacts the efficiency of businesses as well as the lives of employees spending substantial amounts of their time at work. As such, employee satisfaction attracts a lot of attention from researchers. In particular, a lot of effort has been previously devoted to the question of how to positively influence employee satisfaction, for example, through granting benefits. In this paper, we start by empirically exploring a novel dataset comprising two million online employer reviews. Notably, we focus on the analysis of the influencing factors for employee satisfaction. In addition, we leverage our empirical insights to predict employee satisfaction and to assess the predictive strengths of individual factors. We train multiple prediction models and achieve accurate prediction performance (ROC AUC of best model = 0.89 ). We find that the number of benefits received and employment status of reviewers are most predictive, while employee position has less predictive strengths for employee satisfaction. Our work complements existing studies and sheds light on the influencing factors for employee satisfaction expressed in online employer reviews. Employers may use these insights, for example, to correct for biases when assessing their reviews.
AB - Employee satisfaction impacts the efficiency of businesses as well as the lives of employees spending substantial amounts of their time at work. As such, employee satisfaction attracts a lot of attention from researchers. In particular, a lot of effort has been previously devoted to the question of how to positively influence employee satisfaction, for example, through granting benefits. In this paper, we start by empirically exploring a novel dataset comprising two million online employer reviews. Notably, we focus on the analysis of the influencing factors for employee satisfaction. In addition, we leverage our empirical insights to predict employee satisfaction and to assess the predictive strengths of individual factors. We train multiple prediction models and achieve accurate prediction performance (ROC AUC of best model = 0.89 ). We find that the number of benefits received and employment status of reviewers are most predictive, while employee position has less predictive strengths for employee satisfaction. Our work complements existing studies and sheds light on the influencing factors for employee satisfaction expressed in online employer reviews. Employers may use these insights, for example, to correct for biases when assessing their reviews.
KW - Employee satisfaction
KW - Employer reviews
KW - Kununu
UR - http://www.scopus.com/inward/record.url?scp=85093108611&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-60975-7_12
DO - 10.1007/978-3-030-60975-7_12
M3 - Conference paper
SN - 9783030609740
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 152
EP - 167
BT - Social Informatics - 12th International Conference, SocInfo 2020, Proceedings
A2 - Aref, Samin
A2 - Bontcheva, Kalina
A2 - Braghieri, Marco
A2 - Dignum, Frank
A2 - Giannotti, Fosca
A2 - Grisolia, Francesco
A2 - Pedreschi, Dino
T2 - 12th International Conference on Social Informatics
Y2 - 6 October 2020 through 9 October 2020
ER -