데이터 마이닝을 활용한 대학생과 교원의 창업 아이템 분석

The roles of universities include not only education and new knowledge creation, they also include creation of economic value through new technology. The startup items for new firms created by undergraduates and faculties are closely related to the success of those startup companies as the so-called untact businesses increase due to the outbreak of COVID-19. This study analyzes such startup items of undergrads and faculty members who use data mining methodologies based on the R program. To be specific, we simultaneously dealt with 3422 new firms and 615 new firms created by undergraduates and faculties, respectively, from 2020 to 2021 by using three different methodologies, including TF-IDF, a co-occurrence network, and LDA topic modeling. The results found differences in keywords for frequency and importance in both groups. Furthermore, we also found that important keywords and topics were different between startups by undergraduates and those by faculties. Online platforms, fashion items, and online transfer took center positions and were frequent topics in the networks of the former group, whereas remedies, systems, and AI were found to be critical in the networks of the latter group. We suggest various policy implications to increase the probability of success for new firms created by undergrads and faculties.