Data for: STEER-FL: A Process Model for Federated Learning Derived from a Dual-Perspective Systematic Literature Review
Description
Detailed scoring, extraction and evaluation results underlying the STEER-FL process model. The systematic literature review part covers process documentation quality scoring of 55 FL-specific and 29 general ML/DS/DM publications, FL suitability scoring of 19 general process models, and qualitative key aspect extraction from 31 FL-specific publications, all against criteria rated on a three-level ordinal scale (0/1/2). The expert evaluation part documents a semi-structured interview study with eleven domain experts conducted to evaluate and refine the model: the interview structure, five-point Likert ratings on 16 items across five constructs covering technology acceptance, model quality and mechanism efficacy, the qualitative coding of all 263 interview meaning units with theme, statement type (recommendation, positive aspect or note), semantic novelty, cluster references and per-cluster expert counts, the disposition of the 18 recommendations raised independently by at least two experts (incorporated, already covered or deferred), and the resulting per-interview saturation statistics. A README sheet documents the workbook structure and all coding conventions.
Files
Institutions
- Friedrich-Alexander-Universität Erlangen-NürnbergBavaria, Erlangen