We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Möglichkeiten und Grenzen der Optimierung von Verfahren Tiefer Interpretation durch Softwareunterstützung.
- Authors
Schäffer, Burkhard
- Abstract
In this article, parallels are shown between procedures of so-called deep learning in the context of artificial intelligence (AI) and common flow schemata in the field of qualitative-reconstructive procedures of empirical social research: methodologies such as the Documentary Method, Objective Hermeneutics, Grounded Theory or the Narrative Structural Approach all have more or less non-methodisable areas, which are addressed with terms such as abduction or intuition. The parallel to AI is seen in the so-called hidden layers, i.e. the hidden layers of algorithmic computation in language-processing systems, which cannot be reconstructed by computer scientists either, or only with disproportionate effort. Against this background, the aforementioned qualitative-reconstructive methodologies are reconstructed as procedures of deep interpretation and, finally, a research project is presented that investigates to what extent a QDA software for interpreting with the documentary method (DokuMet QDA) can be supported by AI.
- Subjects
ARTIFICIAL intelligence; DEEP learning; COMPUTER scientists; SOCIAL science research; ABDUCTION; INTUITION; RESEARCH methodology
- Publication
Zeitschrift für Qualitative Forschung (ZQF), 2022, Vol 23, Issue 1, p30
- ISSN
2196-2138
- Publication type
Article
- DOI
10.3224/zqf.v23i1.04