This study aims to explore the variations of Korean registers through multi-dimensional analysis by extracting relevant dimensions. These dimensions, derived from the statistical analysis of linguistic features' co-occurrence patterns, play a pivotal role in understanding the linguistic characteristics and functions of various texts and registers. A total of 59 linguistic features were selected to analyze the co-occurrence patterns across 17 traditional registers, leading to the identification of five distinct dimensions. The nature of each dimension was elucidated by examining the contributing linguistic features and the distribution of factor scores across these dimensions. In the final part of the study, the applicability of these dimensions was assessed by analyzing additional registers, such as messenger texts and National Assembly minutes, to determine whether the identified dimensions could serve as an effective framework for analyzing new registers.