Automated Transcription Service: Enabling Transcription at Scale for Social Science Research
Emily Victoria Meanwell, Alan Walsh, Peter van den Berg, Esen Tuna · 2024
Many social scientists collect and analyze qualitative data, often originating in audio recordings drawn from interviews, focus groups, and other face-to-face settings. A longstanding challenge for these researchers has been generating a text transcript of these audio files for use in analysis. Recent advances in artificial intelligence (AI) have transformative potential, but concerns about quality, cost, language support, security, and usability represent a significant barrier to individual researchers. The automated transcription service (ATS) is a completely serverless cloud solution that enables a managed service for social science researchers. ATS provides an easy-to-use service that converts audio in multiple languages into a standard text format with markup to assist in the final editing process.