Extraction Workbook — Trustworthy Knowledge Distillation in Federated Learning: A Systematic Review of Proxy Data Strategies
Abstract
This dataset contains the complete data-extraction workbook of the systematic literature review "Trustworthy Knowledge Distillation in Federated Learning: A Systematic Review of Proxy Data Strategies". The review, conducted under the Kitchenham/Charters guidelines and the PRISMA 2020 reporting standard, covers 51 primary studies (2020–2025) on knowledge distillation in federated learning using synthetic or otherwise non-private proxy data, retrieved from five databases (IEEE Xplore, ACM Digital Library, Scopus, ScienceDirect, and Springer Link). The workbook (.xlsx) is organized as follows: Study_Register — bibliographic and contextual metadata for all 57 screened full-text studies (51 included, 6 excluded with reasons): identifier, title, year, venue, publication type, DOI, final eligibility, application domain, dataset summary, federated-learning setting, heterogeneity type, number of clients, and code/data availability. RQ1–RQ6 sheets — research-question-specific extraction fields for every study, covering proxy use and ownership (RQ1), generation mechanisms and curation policies (RQ2), effects on communication, privacy, and performance (RQ3), proxy-quality criteria and metrics (RQ4), threat models and defenses (RQ5), and methodological gaps and reproducibility (RQ6). Quality_Assessment — per-study scores for the eight-item quality checklist (Q1–Q8; Yes/Partial/No), with computed total scores (mean 6.70, median 7.00, range 4.0–8.0 across the included corpus).…
How strong is the evidence?
Evidence Hub reads this as a systematic review. Its evidence level is 1 of 5 (Highest · pooled evidence). Pooled evidence from several studies sits at the top of the evidence pyramid.
Read the paper
Related studies
- Active Learning during Federated Learning for Object Detection
- Fairness-aware federated learning
- Decentralized Federated Learning
- Building a Personally Identifiable Information Recognizer in a Privacy Preserved Manner Using Automated Annotation and Federated Learning
- Preservation of Patient Level Privacy: Federated Classification and Calibration Models
- Efficient explainable federated learning
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.