Research
Detecting and evaluating fake and AI-generated content — and doing data science rigorously.
My research asks how synthetic and deceptive content can be detected and evaluated as generative AI makes it cheaper and harder to spot. It grew out of my doctoral work on fake reviews and now extends to fake news and AI-generated content across information-systems contexts. A second, ongoing strand concerns the scientific foundations of data science project methodology.
Focus areas
Detection
Fake & AI-generated content
Detecting fake reviews, AI-written text, and synthetic content; feasibility and configuration for practice, including SMEs.
Generative AI
Large language models
How LLMs craft content, output diversity, prompt temperature and text complexity as levers of detectability.
Methodology
Data science process models
Co-author of DASC-PM, a process model for data science and AI projects, and its scientific grounding.
Selected publications
Peer-reviewed, reverse chronological. A complete and up-to-date list is on my ORCID profile and Google Scholar.
- 2026 — Prompt-Temperature and Text Complexity as Determinants of AI-Generated Review Detectability in E-Commerce. SIGSVC Workshop 2025 Proceedings. Link
- 2026 — Controlling Text Complexity for Automatic Text Simplification while Maintaining Semantic Fidelity. SIGSVC Workshop 2025 Proceedings. Link
- 2026 — DASC-PM v2.0 — A Process Model for Data Science and AI Projects (editor). Elmshorn, ISBN 978-3-9824465-3-0. DOI
- 2025 — How to Craft Fake Content Efficiently: Exploring Diversity in Large Language Model Outputs. 58th Hawaii International Conference on System Sciences (HICSS). Link
- 2024 — Battlefield of Online Product Reviews: AI vs. AI. AMCIS 2024 Proceedings. Link
- 2024 — Fake Review Detection on a Budget — Feasibility and Configuration for SME. 28th Pacific Asia Conference on Information Systems (PACIS). Link
- 2023 — Detecting Fake Reviews: Just a Matter of Data. 56th Hawaii International Conference on System Sciences (HICSS). Link
- 2023 — Fake Review Detection — The Value of Domain-Specificity. AMCIS 2023 Proceedings. Link
- 2022 — Vorschlag eines morphologischen Kastens zur Charakterisierung von Data-Science-Projekten. Informatik Spektrum. DOI
- 2021 — Improving Recommender Systems by Using Time-Weighted Sentiment Analysis. 5th International Conference on E-Commerce, E-Business and E-Government. DOI
- 2021 — Where Is the Science in Data Science Projects? (WISDAP). INFORMATIK 2021. DOI
Also co-author of the DASC-PM v1.1 process model and case studies (2022–2023) — see ORCID for the full record.