Using Natural Language Processing to Support Interview Analysis

In social sciences the process of interview analysis is often performed as a form of qualitative data analysis (QDA), to extract relevant information from transcribed interviews. A researcher performing such an analysis is usually assisted by Computer-Assisted Qualitative Data Analysis Software (CAQDAS), which provides a structured framework for coding documents. We expect that this process may significantly benefit from the application of Natural Language Processing (NLP). This thesis provides an overview of potential applications for different NLP technologies. We applied a machine learning approach to automatically generate codings for an existing coding system. During a prototypical implementation of an autocoding algorithm we measured the impact of different NLP techniques on the autocoding process.