As small modular reactors are gaining attention as the future of nuclear energy, it has become increasingly imperative to minimize the number of required operators in order to improve economic competitiveness. Although the tasks of operators have changed and their workloads have been relatively reduced as digital technologies have been applied to advanced main control rooms (MCRs), no change in the number of operators has been made when compared to conventional MCRs in the Republic of Korea. As a solution, the introduction of natural language processing (NLP) technology to replace certain operator tasks in advanced MCRs can be a valuable means of reducing MCR staff. In this paper, we suggest a novel communication framework utilizing NLP technology to minimize the number of operators in advanced MCRs. To do this, we analyze operator tasks in advanced MCRs and select those that can be replaced by NLP technology. We then develop a prototype NLP-based system and analyze the process and characteristics of the suggested communication framework.