Youtube videoları üzerinde Büyük Dil Modeli ile nitel araştırma tema ve kategorileri elde etme


Creative Commons License

AKARSU O. (Yürütücü), YILDIRIM M. S., PARMAKSIZ H.

Yükseköğretim Kurumları Destekli Proje, BAP Araştırma Projesi, 2024 - 2025

  • Proje Türü: Yükseköğretim Kurumları Destekli Proje
  • Destek Programı: BAP Araştırma Projesi
  • Başlama Tarihi: Ekim 2024
  • Bitiş Tarihi: Ekim 2025

Proje Özeti

Qualitative research makes important contributions to understanding social reality, but it is open to criticism due to the subjectivity of researchers and participants in the data collection and interpretation processes. However, qualitative methods are indispensable for revealing deeper meanings that cannot be uncovered using quantitative methods alone in the social sciences. Recent developments in digital technology have brought the use of video data in qualitative research to the forefront. However, new methods are needed for the systematic and objective analysis of big data in video-based research. This project aims to develop an original methodological contribution to the social sciences through the AI-supported qualitative analysis of YouTube videos. In this study, transcripts of videos on entrepreneurial success from the Storybox channel on YouTube were extracted, coded using a large language model, and themes and categories encompassing individual, organizational, and societal levels were identified. The findings show that entrepreneurs' personal characteristics, innovation and collaboration culture at the organizational level, and social capital and cultural values at the societal level play a decisive role in entrepreneurial success. The original value of the project lies in its integration of deep learning models, rarely used in the social sciences, with qualitative analysis. The use of artificial intelligence tools has enabled more systematic and reliable findings by reducing researcher subjectivity. In this respect, the project not only contributes to the entrepreneurship literature but also pioneers the spread of video-based AI-supported analysis in the social sciences.