구조적 토픽 모델을 이용한 상업용 부동산 통계의 제공자와 사용자 간 문제의식 비교

Commercial real estate statistics remain relatively underdeveloped compared with land or housing statistics, limiting their ability to enhance market efficiency. To address this issue, the present study conducted ten rounds of interviews with 33 experts and analyzed the collected data using the Structural Topic Model. The results were as follows. First, statistics providers emphasized the need to reduce the burden of surveys and increase data reliability by sharing data with relevant institutions rather than relying solely on direct surveys. Meanwhile, statistics users highlighted the need to establish statistical areas, such as office regions and retail districts, with greater precision. Both providers and users shared concerns about the sample design and the need to consider a wider range of factors in the design. Second, the Structural Topic Model proved to be useful in analyzing expert interviews, as metadata provided insights into which topics to prioritize to address real-world issues. Although this study confirms the utility of the Structural Topic Model, it does not suggest that the model alone is sufficient for a comprehensive interpretation of interviews. Instead, it recommends the model as a tool that can help researchers interpret interviews more objectively.