Assisted interview transcription

My master’s thesis at the Friedrich-Alexander-Universität Erlangen-Nürnberg.

The Professorship for Open Source Software continuously develops a project named QDAcity, a cloud-based solution for qualitative data analysis (QDA).

Qualitative research deals with a wide array of unstructured input data. One common technique for data gathering is performing interviews which are recorded and subsequently transcribed for analysis. While QDAcity, already supported the analysis stage of this process, the transcription had to be performed manually or with an external tool or service.

Interview media now can be uploaded and an automated transcription is generated.

Transcribing research data involves crucial decisions, like deciding what and how to transcribe (i.e., filler words, pauses). These decisions affect the following analysis and, consequently, the results. Therefore the researcher is often well-advised to transcribe the interview themselves or at least carefully correct the transcription.

A useful user interface for transcription correction supports the correction phase.

I worked with Java on the backend and JS/React on the frontend. The transcription process is done using google cloud platform. Find out more, and the complete document here.