국내 기업 근로자들의 고용브랜드 요인 탐색 및 기업 추천/미추천자 간 차이 분석: 구조적 토픽 모델링 적용

Abstract

This study applies Structural Topic Modeling to explore the text data from corporate reviews of the recruitment platform describing employer brands and to investigate whether such data have meaningful differences between the employees who recommend their firms and those who do not. Specifically, we examine the text data on the satisfactory factors of companies as described by incumbent employees who recommend and do not recommend their firms. Topic modeling derives 30 topics, among which 22 are considered appropriate by using a critical review. These topics are labeled considering previous research on employer brands, and the highest percentage are categorized as ‘horizontal corporate culture’. Significant differences are identified between current employees who recommend and do not recommend their firms. Most of the extracted topics are similar to the variables that are previously used in literature. However, new topics are also derived from our studies. Finally, the implications of study findings are discussed.