Big data industry is an innovative and dynamic working environment based on highly qualified workforce. As the big data phenomenon advances, the demands of the industry for the workforce having these skills and competencies have increased considerably in recent times. Accordingly, the engineering education programs today need to adapt these skills and competencies into their programs. Focusing on this issue, this study aims to extract the core competencies in-demand by the industry. These competencies are the critical ones to better guide the curriculum developers of the engineering education programs. The methodology of the study is based on topic modeling analysis of online job advertisements using Latent Dirichlet Allocation, a generative approach for probabilistic topic models, to automatically discover the trending topics in big data jobs. As a result, domain-specific competencies, developer competencies, soft competencies, business-oriented competencies and analytical competencies are discovered, which revealed that big data competencies contain a wide spectrum of knowledge domains and skill sets based on a multidisciplinary background. The findings of the study are very critical to guide the industry, academia, and big data communities for bridging the gap between the requirements of the industry and the engineering education programs.